Abstract
PostGIS è un'estensione del database object-relational PostgreSQL che consente l'archiviazione di oggetti GIS (Geographic Information Systems). PostGIS comprende il supporto per gli indici spaziali R-Tree basati su GIST, e funzioni per l'analisi e l'elaborazione di oggetti GIS.
Questo è il manuale per la versione 2.3.0dev
Questa opera è rilasciata con una licenza Creative Commons Attribution-Share Alike 3.0 License. Siete liberi di utilizzare questo materiale come volete, ma vi chiediamo di citare il progetto PostGIS come fonte e, ove possibile, di inserire un link a http://postgis.net.
Table of Contents
Table of Contents
PostGIS è stato sviluppato da Refractions Research Inc come progetto di ricerca sui database spaziali. Refractions è una azienda nel settore GIS e database con sede a Victoria, British Columbia, in Canada, specializzata nell'integrazione di dati e nello sviluppo di software su misura. Abbiamo in programma di supportare e sviluppare PostGIS per gestire una gamma di funzionalità GIS importanti, tra cui un completo supporto OpenGIS, operazioni topologiche avanzate (estensioni, superfici, reti), strumenti desktop per la visualizzazione e l'editing di dati GIS, e strumenti web.
PostGIS è in fase di incubazione come progetto della OSGeo Foundation. PostGIS viene continuamente migliorato e sostenuto economicamente da molti sviluppatori FOSS4G e da aziende di tutto il mondo che ricavano grandi benefici dalle sue funzionalità e versatilità.
Il comitato di coordinamento del progetto (in inglese Project Steering Commitee o PSC) coordina la direzione generale, i cicli di rilascio, la documentazione e le iniziative di divulgazione del progetto PostGIS. Inoltre il comitato fornisce supporto agli utenti, accetta e approva patch dalla comunità generale PostGIS e vota su questioni varie che coinvolgono PostGIS come l'accesso di commit per gli sviluppatori, i nuovi membri del comitato e modifiche della API.
Coordina la soluzione dei bug e le iniziative di manutenzione, l'allineamento di PostGIS con i rilasci di PostgreSQL
Manutenzione Buildbot, build sperimentali e di produzione per Windows, documentazione, supporto per gli utenti nel newsgroup PostGIS, supporto X3D, Supporto per il geocodeo Tiger, funzioni di gestione, test di integrità funzionale per nuove funzioni o variazioni sostanziali del codice.
Sviluppo raster, integrazione con GDAL, loader raster, supporto agli utenti, generica soluzione di bachi, prove su vari sistemi operativi (Slackware, Mac, Windows e altri)
Co-fondatore del progetto PostGIS. Correzione dei bug, supporto alla componente geografica, supporto per gli indici geografici e geometrici (2D, 3D, nD e qualsiasi altra cosa legata agli indici spaziali), strutture interne per la geometria, PointCloud (in sviluppo), integrazione delle funzionalità GEOS e allineamento con le nuove versioni di GEOS, loader/dumper, e interfaccia grafica per il loader Shapefile.
Correzione di bug, manutenzione e integrazione delle nuove funzioalità GEOS, allineamento con le versioni di GEOS, supporto topologico, framework Raster e funzioni API di basso livello.
Sviluppo raster, supporto per il driver GDAL, loader
Miglioramenti e aggiunte alla funzione di distanza (comprese le funzioni di distanza e di relazione 3D), formato di output Tiny WKB (TWKB) (in sviluppo) e supporto generale agli utenti
Geometry clustering function additions, other geometry algorithm enhancements, and general user support
Funzioni di input e ouput XML (KML,GML)/GeoJSON, supporto 3D e correzione di bug.
Supporto CMake per PostGIS, ha sviluppato il loader raster originale in python e le funzioni API raster di basso livello.
Architettura complessiva raster, prototipazione, supporto alla programmazione
Sviluppo raster (principalmente funzioni analitiche di map algebra)
Ex-membro del comitato di coordinamento. Sviluppo generale, manutenzione del sito e del buildbot, gestione dell'incubazione OSGeo
Ex-membro del comitato di coordinamento. Documentazione e strumenti di supporto alla documentazione, manutenzione dei buildbot, supporto avanzato per gli utenti nel newsgroup PostGIS, miglioramenti alle funzioni di manutenzione PostGIS.
Lo sviluppatore iniziale e uno dei cofondatori di PostGIS. Dave ha scritto il codice per gli oggetti lato server, il binding degli indici e molte delle funzioni analitiche lato server.
Sviluppo iniziale del loader/dumper per gli Shapefile. Attualmente rappresentante di Project Owner di PostGIS.
Continuo sviluppo e manutenzione delle funzioni di base. Supporto avanzanto per le curve. Interfaccia grafica per il loader di Shapefile.
In alphabetical order: Alex Bodnaru, Alex Mayrhofer, Andrea Peri, Andreas Forø Tollefsen, Andreas Neumann, Anne Ghisla, Barbara Phillipot, Ben Jubb, Bernhard Reiter, Brian Hamlin, Bruce Rindahl, Bruno Wolff III, Bryce L. Nordgren, Carl Anderson, Charlie Savage, Dane Springmeyer, David Skea, David Techer, Eduin Carrillo, Even Rouault, Frank Warmerdam, George Silva, Gerald Fenoy, Gino Lucrezi, Guillaume Lelarge, IIDA Tetsushi, Ingvild Nystuen, Jason Smith, Jeff Adams, Jose Carlos Martinez Llari, Kashif Rasul, Klaus Foerster, Kris Jurka, Leo Hsu, Loic Dachary, Luca S. Percich, Maria Arias de Reyna, Mark Sondheim, Markus Schaber, Maxime Guillaud, Maxime van Noppen, Michael Fuhr, Mike Toews, Nathan Wagner, Nathaniel Clay, Nikita Shulga, Norman Vine, Rafal Magda, Ralph Mason, Richard Greenwood, Silvio Grosso, Steffen Macke, Stephen Frost, Tom van Tilburg, Vincent Mora, Vincent Picavet
Queste sono realtà aziendali o altre istituzioni che hanno contribuito al progetto PostGIS sotto forma di tempo sviluppatore, hosting o finanziamento economico
In alphabetical order: Arrival 3D, Associazione Italiana per l'Informazione Geografica Libera (GFOSS.it), AusVet, Avencia, Azavea, Cadcorp, CampToCamp, CartoDB, City of Boston (DND), Clever Elephant Solutions, Cooperativa Alveo, Deimos Space, Faunalia, Geographic Data BC, Hunter Systems Group, Lidwala Consulting Engineers, LisaSoft, Logical Tracking & Tracing International AG, Michigan Tech Research Institute, Natural Resources Canada, Norwegian Forest and Landscape Institute, Boundless (former OpenGeo), OSGeo, Oslandia, Palantir Technologies, Paragon Corporation, R3 GIS, Refractions Research, Regione Toscana - SITA, Safe Software, Sirius Corporation plc, Stadt Uster, UC Davis Center for Vectorborne Diseases, University of Laval, U.S Department of State (HIU), Zonar Systems
Le campagne di finanziamento diffuso sono campagne che promuoviamo per finanziare funzionalità molto richieste e che possono servire a un gran numero di persone. Ogni campagna è dedicata in modo particolare a una data funzionalità o gruppo di funzionalità. Ogni sponsor partecipa con una quota ridotta rispetto al finanziamento richiesto, e con un numero sufficiente di persone e organizzazioni che contribuiscono, possiamo ottenere i fondi per retribuire un lavoro che servirà a molti. Se avete un'idea per una funzione che pensate possa essere cofinanziata anche da molti altri, vi preghiamo di condividere le vostre idee sul newsgroup PostGIS, e insieme possiamo vedere di realizzarla.
PostGIS 2.0.0 è stata la prima release con cui abbiamo tentato questa strategia. Abbiamo utilizzato PledgeBank, realizzando due campagne di successo.
postgistopology - Oltre 10 sponsor hanno contribuito con 250 USD ciascuno per realizzare la funzione TopoGeometry e per migliorare il supporto della topologia nella versione 2.0.0. E' successo.
postgis64windows - 20 e passa sponsor hanno contribuito con 100 USD ciascuno per retribuire il lavoro necessario per risolvere varie problematiche su PostGIS per Windows a 64 bit. E' successo. Ora abbiamo una versione per PostGIS 2.0.1 disponibile con lo stack builder PostgreSQL.
La libreria per operazioni geometriche GEOS, e il lavoro sugli algoritmi di Martin Davis per far funzionare il tutto, la manutenzione e il supporto continui di Mateusz Loskot, Sandro Santilli (strk), Paul Ramsey a altri.
La libreria GDAL (Geospatial Data Abstraction Library), di Frank Warmerdam e altri viene utilizzata per far funzionare gran parte delle funzionalità raster introdotte in PostGIS 2.0.0. I miglioramenti realizzati su GDAL per consentire il supporto di PostGIS sono stati restituiti come contributo al progetto GDAL.
La libreria di proiezione cartografica Proj4, e il lavoro di Gerald Evenden e Frank Warmerdam nel crearla e mantenerla.
Ultimo, ma non ultimo, il database PostgreSQL, il gigante sulle cui spalle sta PostGIS. Molta della velolcità e flessibilità di PostGIS non sarebbe possibile senza l'estensibilità, il grande query planner, l'indice GIST, e la varietà di caratteristiche SQL rese disponibili da PostgreSQL.
Le versioni più recenti del software, della documentazione e le notizie sono disponibili sul sito web di PostGIS, http://postgis.net.
Ulteriori informazioni sulla libreria di operazioni geometriche GEOS sono disponibili alla pagina http://trac.osgeo.org/geos/.
Ulteriori informazioni sulla libreria di riproiezione Proj4 sono disponibili alla pagina http://trac.osgeo.org/proj/.
Ulteriori informazioni sul server database PostgreSQL database server sono disponibili dal sito principale di PostgreSQL http://www.postgresql.org.
Ulteriori informazioni sull'indicizzazione GiST sono disponibili sul sito di sviluppo di PostgreSQL GiST, http://www.sai.msu.su/~megera/postgres/gist/.
Ulteriori informazioni sul map server internet MapServer sono disponibili su http://mapserver.org.
La documentazione su "Simple Features for Specification for SQL" è disponibile sul sito web dell'OpenGIS Consortium: http://www.opengeospatial.org/.
Table of Contents
Questo capitolo elenca i passi necessari all'installazione di PostGIS.
Per compilare, assumendo di avere tutte le dipendenze nel percorso di ricerca:
tar xvfz postgis-2.3.0dev.tar.gz cd postgis-2.3.0dev ./configure make make install
Una volta installato PostGIS, bisogna abilitarlo individualmente nei database in cui si vuole usare.
Il supporto per i raster è attualmente opzionale, ma installato di default. Per abilitare l'uso di PostGIS come estensione (supportato da PostgreSQL 9.1+) è necessario il supporto raster. Usare PostGIS come estensione è il procedimento preferito e maggiormente user-friendly. Per abilitare spazialmente il tuo database: |
psql -d yourdatabase -c "CREATE EXTENSION postgis;" psql -d yourdatabase -c "CREATE EXTENSION postgis_topology;" -- if you built with sfcgal support -- psql -d yourdatabase -c "CREATE EXTENSION postgis_sfcgal;" -- if you want to install tiger geocoder -- psql -d yourdatabase -c "CREATE EXTENSION fuzzystrmatch" psql -d yourdatabase -c "CREATE EXTENSION postgis_tiger_geocoder;" -- if you installed with pcre -- you should have address standardizer extension as well psql -d yourdatabase -c "CREATE EXTENSION address_standardizer;"
Vedi Section 2.4.3, “Compilazione e installazione delle estensioni PostGIS” per maggiori informazioni riguardo l'ispezione delle estensioni installate/disponibili e il loro aggiornamento, o per passare da un'installazione semplice ad una per estensioni.
For those running who decided for some reason not to compile with raster support, or just are old-fashioned, here are longer more painful instructions for you:
All the .sql files once installed will be installed in share/contrib/postgis-2.2 folder of your PostgreSQL install
createdb yourdatabase createlang plpgsql yourdatabase psql -d yourdatabase -f postgis.sql psql -d yourdatabase -f postgis_comments.sql psql -d yourdatabase -f spatial_ref_sys.sql psql -d yourdatabase -f rtpostgis.sql psql -d yourdatabase -f raster_comments.sql psql -d yourdatabase -f topology.sql psql -d yourdatabase -f topology_comments.sql --if you built with sfcgal support -- psql -d yourdatabase -f sfcgal.sql psql -d yourdatabase -f sfcgal_comments.sql
La restante parte di questo capitolo entra nel dettaglio di ciascuno dei passi di installazione riportati soprta.
As of PostGIS 2.1.3, out-of-db rasters and all raster drivers are disabled by default. In order to re-enable these, you need to set the following environment variables POSTGIS_GDAL_ENABLED_DRIVERS
and POSTGIS_ENABLE_OUTDB_RASTERS
in the server environment. For PostGIS 2.2, you can use the more cross-platform approach of setting the corresponding Section 8.2, “PostGIS Grand Unified Custom Variables (GUCs)”.
If you want to enable offline raster:
POSTGIS_ENABLE_OUTDB_RASTERS=1
Any other setting or no setting at all will disable out of db rasters.
In order to enable all GDAL drivers available in your GDAL install, set this environment variable as follows
POSTGIS_GDAL_ENABLED_DRIVERS=ENABLE_ALL
If you want to only enable specific drivers, set your environment variable as follows:
POSTGIS_GDAL_ENABLED_DRIVERS="GTiff PNG JPEG GIF XYZ"
If you are on windows, do not quote the driver list |
Setting environment variables varies depending on OS. For PostgreSQL installed on Ubuntu or Debian via apt-postgresql, the preferred way is to edit /etc/postgresql/
where 9.3 refers to version of PostgreSQL and main refers to the cluster.9.3
/main
/environment
On windows, if you are running as a service, you can set via System variables which for Windows 7 you can get to by right-clicking on Computer->Properties Advanced System Settings or in explorer navigating to Control Panel\All Control Panel Items\System
. Then clicking Advanced System Settings ->Advanced->Environment Variables and adding new system variables.
After you set the environment variables, you'll need to restart your PostgreSQL service for the changes to take effect.
PostGIS necessita dei seguenti requisiti per la compilazione e l'utilizzo:
Necessari
PostgreSQL 9.1 o versioni superiori. E' necessaria un'installazione completa di PostgreSQL (compresi gli header del server). PostgreSQL è disponibile al sito http://www.postgresql.org .
Per una matrice di supporto completa per PostgreSQL / PostGIS e PostGIS/GEOS vi rimandiamo a http://trac.osgeo.org/postgis/wiki/UsersWikiPostgreSQLPostGIS
Compilatore GNU C (gcc
). Per la compilazione possono essere utilizzati anche altri compilatori ANSI C, ma abbiamo riscontrato molti meno problemi nella compilazione con gcc
.
GNU Make (gmake
or make
). Per molti sistem, GNU make
è la versione di default di make. Potete controllare la versione tramite il comando make -v
. Altre versioni di make
potrebbero non elaborare il Makefile
di PostGIS in modo corretto.
Libreria di riproiezione Proj4, nella versione 4.6.0 o superiore. La libreria Proj4 è utilizzata in PostGIS per fornire supporto alla riproiezione di coordinate. Proj4 è disponibile all'indirizzo http://trac.osgeo.org/proj/ .
GEOS geometry library, version 3.3 or greater, but GEOS 3.5+ is recommended to take full advantage of all the new functions and features. Without GEOS 3.5, you will be missing some major enhancements such as ST_ClipByBox2D and ST_Subdivide. GEOS is available for download from http://trac.osgeo.org/geos/ and 3.4+ is backward-compatible with older versions so fairly safe to upgrade.
LibXML2, versione 2.5.x superiore. LibXML2 è attualmente usata per alcune delle funzioni di importazione (ST_GeomFromGML and ST_GeomFromKML). LibXML2 può essere scaricata da http://xmlsoft.org/downloads.html.
JSON-C, version 0.9 or higher. JSON-C is currently used to import GeoJSON via the function ST_GeomFromGeoJson. JSON-C is available for download from https://github.com/json-c/json-c/releases/.
GDAL, version 1.8 or higher (1.9 or higher is strongly recommended since some things will not work well or behavior differently with lower versions). This is required for raster support and to be able to install with CREATE EXTENSION postgis
so highly recommended for those running 9.1+. http://trac.osgeo.org/gdal/wiki/DownloadSource.
Opzionali
GDAL (pseudo optional) only if you don't want raster and don't care about installing with CREATE EXTENSION postgis
can you leave it out. Keep in mind other extensions may have a requires postgis extension which will prevent you from installing them unless you install postgis as an extension. So it is highly recommended you compile with GDAL support.
Also make sure to enable the drivers you want to use as described in Section 2.1, “Versione sintetica”.
GTK (è necessario GTK+2.0, 2.8+) serve per compilare il loader shp2pgsql-gui shape. http://www.gtk.org/ .
SFCGAL, version 1.1 (or higher) could be used to provide additional 2D and 3D advanced analysis functions to PostGIS cf Section 8.10, “SFCGAL Functions”. And also allow to use SFCGAL rather than GEOS for some 2D functions provided by both backends (like ST_Intersection or ST_Area, for instance). A PostgreSQL configuration variable postgis.backend
allow end user to control which backend he want to use if SFCGAL is installed (GEOS by default). Nota: SFCGAL 1.2 require at least CGAL 4.3 and Boost 1.54 (cf: http://oslandia.github.io/SFCGAL/installation.html) https://github.com/Oslandia/SFCGAL.
In order to build the Chapter 12, Address Standardizer you will also need PCRE http://www.pcre.org (which generally is already installed on nix systems). Regex::Assemble
perl CPAN package is only needed if you want to rebuild the data encoded in parseaddress-stcities.h
. Chapter 12, Address Standardizer will automatically be built if it detects a PCRE library, or you pass in a valid --with-pcre-dir=/path/to/pcre
during configure.
CUnit (CUnit
). Serve per i collaudi di regressione. http://cunit.sourceforge.net/
DocBook (xsltproc
) è necessario per creare la documentazione. Docbook è disponibile sul sito http://www.docbook.org/ .
DBLatex (dblatex
) è necessario per creare la documentazione in formato PDF. DBLatex è disponibile dal sito http://dblatex.sourceforge.net/ .
ImageMagick (convert
) viene utilizzato per generare le immagini utilizzate nella documentazione. ImageMagick è disponibile sul sito http://www.imagemagick.org/ .
Retrieve the PostGIS source archive from the downloads website http://postgis.net/stuff/postgis-2.3.0dev.tar.gz
wget http://postgis.net/stuff/postgis-2.3.0dev.tar.gz tar -xvzf postgis-2.3.0dev.tar.gz
Questo passaggio creerà una cartella denominata postgis-2.3.0dev
nella cartella di lavoro attuale.
In alternativa, potete eseguire un checkout del sorgente dalla repository svn http://svn.osgeo.org/postgis/trunk/ .
svn checkout http://svn.osgeo.org/postgis/trunk/ postgis-2.3.0dev
Entrate nella cartella appena creata, postgis-2.3.0dev
, per proseguire con l'installazione.
Ormai molti sistemi operativi contengono pacchetti precompilati per PostgreSQL/PostGIS. In molti casi la compilazione è necessaria solamente se volete l'ultimissima versione o se siete uno dei manutentori dei pacchetti. Questo paragrafo comprende le istruzioni generali per la compilazione. Se state compilando su Windows o su un altro sistema operativo, potete trovare istruzioni più dettagliate alle pagine PostGIS User contributed compile guides e PostGIS Dev Wiki. I pacchetti precompilati per i vari sistemi operativi sono elencati alla pagina PostGIS Pre-built Packages Se siete un utente Windows, potete reperire le versioni stabili tramite Stackbuilder o la pagina PostGIS Windows download. Inoltre potete scaricare le ultimissime versioni sperimentati per Windows, che vengono create una o due volte alla settimana, quando ci sono aggiornamenti interessanti. Potete utilizzare queste versioni per fare prove con le versioni in via di sviluppo di PostGIS |
Il modulo PostGIS è un'estensione del server PostgreSQL. Pertanto, + indispensabile che PostGIS 2.3.0dev abbia pieno accesso agli header di PostgreSQL server per potersi compilare. Può essere compilato con versioni di PostgreSQL dalla 9.1 in su. Le versioni precedenti di PostgreSQL non sono supportate.
Vi rimandiamo alle guide per l'installazione di PostgreSQL, se questo non è stato ancora installato. http://www.postgresql.org .
Per le funzionalità legate a GEOS, quando installate PostgresSQL è possibile che dobbiate esplicitamente linkare PostgreSQL con la libreria C++ standard: LDFLAGS=-lstdc++ ./configure [INSERITE QUI LE VOSTRE OPZIONI] Questa è una soluzione alla buona per l'interazione con finte eccezioni C++ con gli strumenti di sviluppo più datati. Se riscontrate problemi anomali (chiusura inattesa del server o casi simili), provate questo trucco. Ciò richiederà ovviamente di ricompilare PostgreSQL da zero. |
I passaggi seguenti ripercorrono la procedura per configurare e compilare il sorgente di PostGIS. Sono scritti per utenti Linux e non funzioneranno su Windows o Mac.
Come in molte installazioni Linux, il primo passo consiste nel generare il Makefile che sarà poi utilizzato che compilare il codice sorgente. Per questo si esegue lo script da shell
./configure
Se non si specificano parametri aggiuntivi, questo comando cercherà di individuare automaticamente le componenti e le librerie necessarie per compilare il codice sorgente PostGIS sul vostro sistema. Benché questa sia la modalità di utilizzo più comune per ./configure, lo script accetta vari parametri per coloro che tengono le librerie e i programmi necessari in percorsi non standard.
La lista seguente riporta solamente i parametri usati più di frequente. Per una lista completa, utilizzate i parametri --help o --help=short.
Questo è il percorso in cui saranno installate le librerie PostGIS e gli script SQL. Come default, questo percorso coincide con quello rilevato per l'installazione di PostgreSQL.
Attualmente questo parametro non funzione, dato che il pacchetto si installerà solamente nella cartella di installazione di PostgreSQL. Potete visitare la pagina http://trac.osgeo.org/postgis/ticket/635 per seguire l'evoluzione di questo bug. |
PostgreSQL fornisce una utility chiamata pg_config per consentire a estensioni quali PostGIS di individuare la cartella di installazione di PostgreSQL. Utilizzate questo parametro (--with-pgconfig=/path/to/pg_config) per specificare manualmente una particolare installazione di PostgreSQL per la quale PostGIS sarà compilato.
GDAL è una libreria necessaria per il supporto raster. Utilizzare il comando gdal-config per consentire all'installazione di individuare la cartella di installazione di GDAL, oppure utilizzare il parametro --with-gdalconfig=/path/to/gdal-config per indicare manualmente specifica installazione di GDAL con cui compilare PostGIS.
GEOS, una libreria necessaria per gestire le geometrie, fornisce una utility chiamata geos-config per consentire alla procedura di installazione di individuare la cartella di installazione di GEOS. E' possibile utilizzare questo parametro (--with-geosconfig=/path/to/geos-config) per indicare manualmente una specifica installazione di GEO da utilizzare per la compilazione.
LibXML è la libreria richiesta dai comandi GeomFromKML/GML. Se la libreria è installata, viene normalmente localizzata automaticamente, ma se non è installata, o se volete utilizzare una versione specifica, dovete rimandare PostGIS a uno specifico file di configurazione xml2-config
per consentire all'installazione di localizzare la cartella di installazione di LibXML. Utilizzate questo parametro (
>--with-xml2config=/path/to/xml2-config) per indicare manualmente il riferimento a una specifica installazione di LibXML per la compilazione di PostGIS.
Proj4 è una libreria di riproiezione richiesta da PostGIS. Utilizzate questo parametro (--with-projdir=/path/to/projdir) per indicare manualmente una specifica cartella di installazione di Proj4 per la compilazione di PostGIS.
Cartella di installazione di iconv.
JSON-C è una libreria JSON con licenza MIR utilizzata da PostGIS per supportare ST_GeomFromJSON. Utilizzare questo parametro (--with-jsondir=/path/to/jsondir) per indicare manualmente una specifica cartella di installazione che PostGIS userà per la compilazione.
PCRE is an BSD-licensed Perl Compatible Regular Expression library required by address_standardizer extension. Use this parameter (--with-pcredir=/path/to/pcredir) to manually specify a particular PCRE installation directory that PostGIS will build against.
Compilare l'interfaccia grafica per l'importazione di dati (richiede GTK+2.0). Questo passaggio creerà shp2pgsql-gui, interfaccia grafica per shp2pgsql.
Compilare con supporto raster. Questo compilerà la libreria rtpostgis-2.3.0dev e il file rtpostgis.sql. Potrebbe non essere necessario nella release finale, dato che si prevede di avere il supporto raster come default.
Compilare con il supporto per la topologia. Questo creerà il file topology.sql. Non c'è una libreria corrispondente, in quanto tutta la logica per la topologia è contenuta nella libreria postgis-2.3.0dev.
Normalmente PostGIS cercherà di rilevare il supporto per gettext e di utilizzarlo nella compilazione. Tuttavia, se risocontrate problemi di compatibilità che determinano interruzioni del loader, potete disabilitare il supporto con questo comando. Potete fare riferimento al ticket http://trac.osgeo.org/postgis/ticket/748 per un esempio di un caso risolto utilizzando questa particolare configurazione. NOTA: disattivanto questa opzione non vi perdete molto, dato che è utilizzata per il supporto internazionale o per le etichette nel loader, che non sono ancora ben documentate e in forma sperimentale.
By default PostGIS will not install with sfcgal support without this switch. PATH
is an optional argument that allows to specify an alternate PATH to sfcgal-config.
Se avete ottenuto PostGIS dalla repository SVN , il primo passo consiste nell'eseguire lo script ./autogen.sh Questo script genererà lo script configure, che a sua volta viente utilizzato per personalizzare l'installazione di PostGIS. Se invece avete ottenuto PostGIS come file tar, non sarà necessario eseguire ./autogen.sh, dato che configure sarà già stato generato. |
Una volta creato il Makefile, compilare PostGIS è semplice come eseguire
make
L'ultima linea dei messaggi in uscita dovrebbe essere "PostGIS was built successfully. Ready to install.
"
A partire PostGIS v1.4.0, tutte le funzioni hanno commenti generati dalla documentazione. Se desiderate installare questi commenti nei vostri database spaziali in seguito, eseguite il comando che richiede docbook. Il file postgis_comments.sql e altri file di commenti per i pacchetti, raster_comments.sql e topology_comments.sql, sono presenti nella cartella doc della distribuzione tar.gz. Pertanto non è necessario di eseguire il make dei commenti se si esegue l'installazione a partire dal file tar.
make comments
Introdotto in PostGIS 2.0. Questa opzione genera degli opuscoli riassuntivi in formato HTML, utili come riferimento speditivo o per distribuzione agli studenti. Richiede xsltproc e genererà 4 file della cartella doc: topology_cheatsheet.html
, tiger_geocoder_cheatsheet.html
, raster_cheatsheet.html
, postgis_cheatsheet.html
Ne potete scaricare alcuni già pronti sia in formato HTML che PDF dalla pagina PostGIS / PostgreSQL Study Guides
make cheatsheets
Le estensioni di PostGIS vengono compilate e installate automaticamente se state utilizzando PostgreSQL 9.1+.
Se state compilando a partire dalla repository del sorgente, dovete prima compilare le descrizioni delle funzioni. Queste vengono create se avete docbook installato. Potete anche eseguire la compilazione manualmente tramite il comando:
make comments
Creare i file dei commenti non è necessario se state compilando a partire dal file tar, dato che questi file sono inclusi nel file tar stesso.
Se state compilando per PostgreSQL 9.1, le estensioni dovrebbero essere create automaticamente nel corso del processo di make install. Se serve, potete eseguire la compilazione a partire dalle cartelle delle estensioni, o potete copiare i file se ne avete bisogno su un altro server.
cd extensions cd postgis make clean make make install cd .. cd postgis_topology make clean make make install cd .. cd postgis_sfcgal make clean make make install cd .. cd address_standardizer make clean make make install make installcheck cd .. cd postgis_tiger_geocoder make clean make make install make installcheck
I file delle estensioni sono sempre gli stessi a pari versione di PostGIS, a prescindere dal sistema operativo, per cui non ci sono problemi a copiare i file delle estensioni da un sistema operativo a un altro, fintanto che i file binari di PostGIS sono già installati sui vostri server.
Se volete installare le estensioni manualmente su un server separato, diverso da quello di sviluppo, dovete copiare i seguenti file dalla cartella delle estensioni nella cartella PostgreSQL / share / extension
della vostra installazione PostgreSQL, così come i file binari per la versione normale di PostGIS, se questi già non sono presenti sul server.
Questi sono i file di controllo che indicano informazioni come la versione dell'estensione da installare, se non specificata. postgis.control, postgis_topology.control
.
Tutti i file nella cartella /sql di ciascuna estesnione. Da notare che quesi devono essere copiati nella cartella principale delle estensioni di PostgreSQL, share/extension extensions/postgis/sql/*.sql
, extensions/postgis_topology/sql/*.sql
Una volta eseguita questa operazione, dovreste vedere postgis
, postgis_topology
come estensioni disponibili in PgAdmin -> extensions.
Se state utilizzando psql, potete verificare l'avvenuta installazione delle estensioni tramite questa query:
SELECT name, default_version,installed_version FROM pg_available_extensions WHERE name LIKE 'postgis%' or name LIKE 'address%'; name | default_version | installed_version ------------------------------+-----------------+------------------- address_standardizer | 2.3.0dev | 2.3.0dev address_standardizer_data_us | 2.3.0dev | 2.3.0dev postgis | 2.3.0dev | 2.3.0dev postgis_sfcgal | 2.3.0dev | postgis_tiger_geocoder | 2.3.0dev | 2.3.0dev postgis_topology | 2.3.0dev | (6 rows)
Se avete le estensioni installate nel database che state interrogando, vedrete questo fatto menzionato nella colonna installed_version
. Se non vengono riportari record nel risultato, significa che non avete alcuna estensione installata sul server. PgAdmin III 1.14+ fornirà inoltre questa informazione nella sezione extensions
dell'albero di navigazione del database e consentirà anche l'aggiornamento o la disinstallazione tramite clic del tasto destro del mouse.
Se le estensioni sono disponibili, potete installare l'estensione postgis nel database da voi scelto utilizzando l'interfaccia pgAdmin per le estensioni o eseguendo i seguenti comandi SQL:
CREATE EXTENSION postgis; CREATE EXTENSION postgis_sfcgal; CREATE EXTENSION fuzzystrmatch; --needed for postgis_tiger_geocoder --optional used by postgis_tiger_geocoder, or can be used standalone CREATE EXTENSION address_standardizer; CREATE EXTENSION address_standardizer_data_us; CREATE EXTENSION postgis_tiger_geocoder; CREATE EXTENSION postgis_topology;
In psql you can use to see what versions you have installed and also what schema they are installed.
\connect mygisdb \x \dx postgis*
List of installed extensions -[ RECORD 1 ]------------------------------------------------- - Name | postgis Version | 2.3.0dev Schema | public Description | PostGIS geometry, geography, and raster spat.. -[ RECORD 2 ]------------------------------------------------- - Name | postgis_tiger_geocoder Version | 2.3.0dev Schema | tiger Description | PostGIS tiger geocoder and reverse geocoder -[ RECORD 3 ]------------------------------------------------- - Name | postgis_topology Version | 2.3.0dev Schema | topology Description | PostGIS topology spatial types and functions
Per le tabelle delle estensioni |
If you installed 2.3.0dev, without using our wonderful extension system, you can change it to be extension based by first upgrading to the latest micro version running the upgrade scripts: postgis_upgrade_22_minor.sql
,raster_upgrade_22_minor.sql
,topology_upgrade_22_minor.sql
.
Se avete installato PostGIS senza supporto raster, dovrete prima installare il supporto raster (utilizzando tutto il file rtpostgis.sql
Poi potete eseguire i comandi seguenti per creare i pacchetti di funzioni nelle estensioni corrispondenti.
CREATE EXTENSION postgis FROM unpackaged; CREATE EXTENSION postgis_topology FROM unpackaged; CREATE EXTENSION postgis_tiger_geocoder FROM unpackaged;
Se volete eseguire un test sul PostGIS compilato, eseguite
make check
Il comando di cui sopra eseguirà vari controlli e collaudi di regressione, utilizzando la libreria generata per un database PostgreSQL effettivo.
Se avete configurato PostGIS utilizzando percorsi non standard per PostgreSQL, GEOS o Proj4, potreste dover aggiungere il percorso di queste librerie nella variabile di ambiente LD_LIBRARY_PATH |
Attualmente il comando make check fa riferimento alle variabili di ambiente |
Se il test è positivo, l'uscita dovrebbe essere simile alla seguente:
CUnit - A unit testing framework for C - Version 2.1-2 http://cunit.sourceforge.net/ Suite: computational_geometry Test: test_lw_segment_side ...passed Test: test_lw_segment_intersects ...passed Test: test_lwline_crossing_short_lines ...passed Test: test_lwline_crossing_long_lines ...passed Test: test_lwline_crossing_bugs ...passed Test: test_lwpoint_set_ordinate ...passed Test: test_lwpoint_get_ordinate ...passed Test: test_point_interpolate ...passed Test: test_lwline_clip ...passed Test: test_lwline_clip_big ...passed Test: test_lwmline_clip ...passed Test: test_geohash_point ...passed Test: test_geohash_precision ...passed Test: test_geohash ...passed Test: test_geohash_point_as_int ...passed Test: test_isclosed ...passed Test: test_lwgeom_simplify ...passed Suite: buildarea Test: buildarea1 ...passed Test: buildarea2 ...passed Test: buildarea3 ...passed Test: buildarea4 ...passed Test: buildarea4b ...passed Test: buildarea5 ...passed Test: buildarea6 ...passed Test: buildarea7 ...passed Suite: geometry_clean Test: test_lwgeom_make_valid ...passed Suite: clip_by_rectangle Test: test_lwgeom_clip_by_rect ...passed Suite: force_sfs Test: test_sfs_11 ...passed Test: test_sfs_12 ...passed Test: test_sqlmm ...passed Suite: geodetic Test: test_sphere_direction ...passed Test: test_sphere_project ...passed Test: test_lwgeom_area_sphere ...passed Test: test_signum ...passed Test: test_gbox_from_spherical_coordinates ...passed Test: test_gserialized_get_gbox_geocentric ...passed Test: test_clairaut ...passed Test: test_edge_intersection ...passed Test: test_edge_intersects ...passed Test: test_edge_distance_to_point ...passed Test: test_edge_distance_to_edge ...passed Test: test_lwgeom_distance_sphere ...passed Test: test_lwgeom_check_geodetic ...passed Test: test_gserialized_from_lwgeom ...passed Test: test_spheroid_distance ...passed Test: test_spheroid_area ...passed Test: test_lwpoly_covers_point2d ...passed Test: test_gbox_utils ...passed Test: test_vector_angle ...passed Test: test_vector_rotate ...passed Test: test_lwgeom_segmentize_sphere ...passed Test: test_ptarray_contains_point_sphere ...passed Test: test_ptarray_contains_point_sphere_iowa ...passed Suite: GEOS Test: test_geos_noop ...passed Test: test_geos_subdivide ...passed Test: test_geos_linemerge ...passed Suite: Clustering Test: basic_test ...passed Test: nonsequential_test ...passed Test: basic_distance_test ...passed Test: single_input_test ...passed Test: empty_inputs_test ...passed Suite: Clustering Union-Find Test: test_unionfind_create ...passed Test: test_unionfind_union ...passed Test: test_unionfind_ordered_by_cluster ...passed Suite: homogenize Test: test_coll_point ...passed Test: test_coll_line ...passed Test: test_coll_poly ...passed Test: test_coll_coll ...passed Test: test_geom ...passed Test: test_coll_curve ...passed Suite: encoded_polyline_input Test: in_encoded_polyline_test_geoms ...passed Test: in_encoded_polyline_test_precision ...passed Suite: geojson_input Test: in_geojson_test_srid ...passed Test: in_geojson_test_bbox ...passed Test: in_geojson_test_geoms ...passed Suite: twkb_input Test: test_twkb_in_point ...passed Test: test_twkb_in_linestring ...passed Test: test_twkb_in_polygon ...passed Test: test_twkb_in_multipoint ...passed Test: test_twkb_in_multilinestring ...passed Test: test_twkb_in_multipolygon ...passed Test: test_twkb_in_collection ...passed Test: test_twkb_in_precision ...passed Suite: serialization/deserialization Test: test_typmod_macros ...passed Test: test_flags_macros ...passed Test: test_serialized_srid ...passed Test: test_gserialized_from_lwgeom_size ...passed Test: test_gbox_serialized_size ...passed Test: test_lwgeom_from_gserialized ...passed Test: test_lwgeom_count_vertices ...passed Test: test_on_gser_lwgeom_count_vertices ...passed Test: test_geometry_type_from_string ...passed Test: test_lwcollection_extract ...passed Test: test_lwgeom_free ...passed Test: test_lwgeom_flip_coordinates ...passed Test: test_f2d ...passed Test: test_lwgeom_clone ...passed Test: test_lwgeom_force_clockwise ...passed Test: test_lwgeom_calculate_gbox ...passed Test: test_lwgeom_is_empty ...passed Test: test_lwgeom_same ...passed Test: test_lwline_from_lwmpoint ...passed Test: test_lwgeom_as_curve ...passed Test: test_lwgeom_scale ...passed Test: test_gserialized_is_empty ...passed Test: test_gbox_same_2d ...passed Suite: measures Test: test_mindistance2d_tolerance ...passed Test: test_rect_tree_contains_point ...passed Test: test_rect_tree_intersects_tree ...passed Test: test_lwgeom_segmentize2d ...passed Test: test_lwgeom_locate_along ...passed Test: test_lw_dist2d_pt_arc ...passed Test: test_lw_dist2d_seg_arc ...passed Test: test_lw_dist2d_arc_arc ...passed Test: test_lw_arc_length ...passed Test: test_lw_dist2d_pt_ptarrayarc ...passed Test: test_lw_dist2d_ptarray_ptarrayarc ...passed Test: test_lwgeom_tcpa ...passed Test: test_lwgeom_is_trajectory ...passed Suite: effectivearea Test: do_test_lwgeom_effectivearea_lines ...passed Test: do_test_lwgeom_effectivearea_polys ...passed Suite: miscellaneous Test: test_misc_force_2d ...passed Test: test_misc_simplify ...passed Test: test_misc_count_vertices ...passed Test: test_misc_area ...passed Test: test_misc_wkb ...passed Test: test_grid ...passed Suite: noding Test: test_lwgeom_node ...passed Suite: encoded_polyline_output Test: out_encoded_polyline_test_geoms ...passed Test: out_encoded_polyline_test_srid ...passed Test: out_encoded_polyline_test_precision ...passed Suite: geojson_output Test: out_geojson_test_precision ...passed Test: out_geojson_test_dims ...passed Test: out_geojson_test_srid ...passed Test: out_geojson_test_bbox ...passed Test: out_geojson_test_geoms ...passed Suite: gml_output Test: out_gml_test_precision ...passed Test: out_gml_test_srid ...passed Test: out_gml_test_dims ...passed Test: out_gml_test_geodetic ...passed Test: out_gml_test_geoms ...passed Test: out_gml_test_geoms_prefix ...passed Test: out_gml_test_geoms_nodims ...passed Test: out_gml2_extent ...passed Test: out_gml3_extent ...passed Suite: kml_output Test: out_kml_test_precision ...passed Test: out_kml_test_dims ...passed Test: out_kml_test_geoms ...passed Test: out_kml_test_prefix ...passed Suite: svg_output Test: out_svg_test_precision ...passed Test: out_svg_test_dims ...passed Test: out_svg_test_relative ...passed Test: out_svg_test_geoms ...passed Test: out_svg_test_srid ...passed Suite: x3d_output Test: out_x3d3_test_precision ...passed Test: out_x3d3_test_geoms ...passed Test: out_x3d3_test_option ...passed Suite: ptarray Test: test_ptarray_append_point ...passed Test: test_ptarray_append_ptarray ...passed Test: test_ptarray_locate_point ...passed Test: test_ptarray_isccw ...passed Test: test_ptarray_signed_area ...passed Test: test_ptarray_unstroke ...passed Test: test_ptarray_insert_point ...passed Test: test_ptarray_contains_point ...passed Test: test_ptarrayarc_contains_point ...passed Test: test_ptarray_scale ...passed Suite: printing Test: test_lwprint_default_format ...passed Test: test_lwprint_format_orders ...passed Test: test_lwprint_optional_format ...passed Test: test_lwprint_oddball_formats ...passed Test: test_lwprint_bad_formats ...passed Suite: SFCGAL Test: test_sfcgal_noop ...passed Suite: split Test: test_lwline_split_by_point_to ...passed Test: test_lwgeom_split ...passed Suite: stringbuffer Test: test_stringbuffer_append ...passed Test: test_stringbuffer_aprintf ...passed Suite: surface Test: triangle_parse ...passed Test: tin_parse ...passed Test: polyhedralsurface_parse ...passed Test: surface_dimension ...passed Suite: Internal Spatial Trees Test: test_tree_circ_create ...passed Test: test_tree_circ_pip ...passed Test: test_tree_circ_pip2 ...passed Test: test_tree_circ_distance ...passed Test: test_tree_circ_distance_threshold ...passed Suite: triangulate Test: test_lwgeom_delaunay_triangulation ...passed Suite: twkb_output Test: test_twkb_out_point ...passed Test: test_twkb_out_linestring ...passed Test: test_twkb_out_polygon ...passed Test: test_twkb_out_multipoint ...passed Test: test_twkb_out_multilinestring ...passed Test: test_twkb_out_multipolygon ...passed Test: test_twkb_out_collection ...passed Test: test_twkb_out_idlist ...passed Suite: varint Test: test_zigzag ...passed Test: test_varint ...passed Test: test_varint_roundtrip ...passed Suite: wkb_input Test: test_wkb_in_point ...passed Test: test_wkb_in_linestring ...passed Test: test_wkb_in_polygon ...passed Test: test_wkb_in_multipoint ...passed Test: test_wkb_in_multilinestring ...passed Test: test_wkb_in_multipolygon ...passed Test: test_wkb_in_collection ...passed Test: test_wkb_in_circularstring ...passed Test: test_wkb_in_compoundcurve ...passed Test: test_wkb_in_curvpolygon ...passed Test: test_wkb_in_multicurve ...passed Test: test_wkb_in_multisurface ...passed Test: test_wkb_in_malformed ...passed Suite: wkb_output Test: test_wkb_out_point ...passed Test: test_wkb_out_linestring ...passed Test: test_wkb_out_polygon ...passed Test: test_wkb_out_multipoint ...passed Test: test_wkb_out_multilinestring ...passed Test: test_wkb_out_multipolygon ...passed Test: test_wkb_out_collection ...passed Test: test_wkb_out_circularstring ...passed Test: test_wkb_out_compoundcurve ...passed Test: test_wkb_out_curvpolygon ...passed Test: test_wkb_out_multicurve ...passed Test: test_wkb_out_multisurface ...passed Test: test_wkb_out_polyhedralsurface ...passed Suite: wkt_input Test: test_wkt_in_point ...passed Test: test_wkt_in_linestring ...passed Test: test_wkt_in_polygon ...passed Test: test_wkt_in_multipoint ...passed Test: test_wkt_in_multilinestring ...passed Test: test_wkt_in_multipolygon ...passed Test: test_wkt_in_collection ...passed Test: test_wkt_in_circularstring ...passed Test: test_wkt_in_compoundcurve ...passed Test: test_wkt_in_curvpolygon ...passed Test: test_wkt_in_multicurve ...passed Test: test_wkt_in_multisurface ...passed Test: test_wkt_in_tin ...passed Test: test_wkt_in_polyhedralsurface ...passed Test: test_wkt_in_errlocation ...passed Suite: wkt_output Test: test_wkt_out_point ...passed Test: test_wkt_out_linestring ...passed Test: test_wkt_out_polygon ...passed Test: test_wkt_out_multipoint ...passed Test: test_wkt_out_multilinestring ...passed Test: test_wkt_out_multipolygon ...passed Test: test_wkt_out_collection ...passed Test: test_wkt_out_circularstring ...passed Test: test_wkt_out_compoundcurve ...passed Test: test_wkt_out_curvpolygon ...passed Test: test_wkt_out_multicurve ...passed Test: test_wkt_out_multisurface ...passed Run Summary: Type Total Ran Passed Failed Inactive suites 38 38 n/a 0 0 tests 251 251 251 0 0 asserts 2468 2468 2468 0 n/a Elapsed time = 0.298 seconds Creating database 'postgis_reg' Loading PostGIS into 'postgis_reg' /projects/postgis/branches/2.2/regress/00-regress-install/share/contrib/postgis/postgis.sql /projects/postgis/branches/2.2/regress/00-regress-install/share/contrib/postgis/postgis_comments.sql Loading SFCGAL into 'postgis_reg' /projects/postgis/branches/2.2/regress/00-regress-install/share/contrib/postgis/sfcgal.sql /projects/postgis/branches/2.2/regress/00-regress-install/share/contrib/postgis/sfcgal_comments.sql PostgreSQL 9.4.4, compiled by Visual C++ build 1800, 32-bit Postgis 2.2.0dev - r13980 - 2015-08-23 06:13:07 scripts 2.2.0dev r13980 GEOS: 3.5.0-CAPI-1.9.0 r4088 PROJ: Rel. 4.9.1, 04 March 2015 SFCGAL: 1.1.0 Running tests loader/Point .............. ok loader/PointM .............. ok loader/PointZ .............. ok loader/MultiPoint .............. ok loader/MultiPointM .............. ok loader/MultiPointZ .............. ok loader/Arc .............. ok loader/ArcM .............. ok loader/ArcZ .............. ok loader/Polygon .............. ok loader/PolygonM .............. ok loader/PolygonZ .............. ok loader/TSTPolygon ......... ok loader/TSIPolygon ......... ok loader/TSTIPolygon ......... ok loader/PointWithSchema ..... ok loader/NoTransPoint ......... ok loader/NotReallyMultiPoint ......... ok loader/MultiToSinglePoint ......... ok loader/ReprojectPts ........ ok loader/ReprojectPtsGeog ........ ok loader/Latin1 .... ok loader/Latin1-implicit .... ok loader/mfile .... ok dumper/literalsrid ....... ok dumper/realtable ....... ok affine .. ok bestsrid .. ok binary .. ok boundary .. ok cluster .. ok concave_hull .. ok ctors .. ok dump .. ok dumppoints .. ok empty .. ok forcecurve .. ok geography .. ok in_geohash .. ok in_gml .. ok in_kml .. ok in_encodedpolyline .. ok iscollection .. ok legacy .. ok long_xact .. ok lwgeom_regress .. ok measures .. ok operators .. ok out_geometry .. ok out_geography .. ok polygonize .. ok polyhedralsurface .. ok postgis_type_name .. ok regress .. ok regress_bdpoly .. ok regress_index .. ok regress_index_nulls .. ok regress_management .. ok regress_selectivity .. ok regress_lrs .. ok regress_ogc .. ok regress_ogc_cover .. ok regress_ogc_prep .. ok regress_proj .. ok relate .. ok remove_repeated_points .. ok removepoint .. ok setpoint .. ok simplify .. ok simplifyvw .. ok size .. ok snaptogrid .. ok split .. ok sql-mm-serialize .. ok sql-mm-circularstring .. ok sql-mm-compoundcurve .. ok sql-mm-curvepoly .. ok sql-mm-general .. ok sql-mm-multicurve .. ok sql-mm-multisurface .. ok swapordinates .. ok summary .. ok temporal .. ok tickets .. ok twkb .. ok typmod .. ok wkb .. ok wkt .. ok wmsservers .. ok knn .. ok hausdorff .. ok regress_buffer_params .. ok offsetcurve .. ok relatematch .. ok isvaliddetail .. ok sharedpaths .. ok snap .. ok node .. ok unaryunion .. ok clean .. ok relate_bnr .. ok delaunaytriangles .. ok clipbybox2d .. ok subdivide .. ok in_geojson .. ok regress_sfcgal .. ok sfcgal/empty .. ok sfcgal/geography .. ok sfcgal/legacy .. ok sfcgal/measures .. ok sfcgal/regress_ogc_prep .. ok sfcgal/regress_ogc .. ok sfcgal/regress .. ok sfcgal/tickets .. ok sfcgal/concave_hull .. ok sfcgal/wmsservers .. ok sfcgal/approximatemedialaxis .. ok uninstall . /projects/postgis/branches/2.2/regress/00-regress-install/share/contrib/postgis/uninstall_sfcgal.sql /projects/postgis/branches/2.2/regress/00-regress-install/share/contrib/postgis/uninstall_postgis.sql . ok (4336) Run tests: 118 Failed: 0 -- if you built --with-gui, you should see this too CUnit - A unit testing framework for C - Version 2.1-2 http://cunit.sourceforge.net/ Suite: Shapefile Loader File shp2pgsql Test Test: test_ShpLoaderCreate() ...passed Test: test_ShpLoaderDestroy() ...passed Suite: Shapefile Loader File pgsql2shp Test Test: test_ShpDumperCreate() ...passed Test: test_ShpDumperDestroy() ...passed Run Summary: Type Total Ran Passed Failed Inactive suites 2 2 n/a 0 0 tests 4 4 4 0 0 asserts 4 4 4 0 n/a
The postgis_tiger_geocoder
and address_standardizer
extensions, currenlty only support the standard PostgreSQL installcheck. To test these use the below. Note: the make install is not necessary if you already did make install at root of PostGIS code folder.
For address_standardizer:
cd extensions/address_standardizer make install make installcheck
Output should look like:
============== dropping database "contrib_regression" ============== DROP DATABASE ============== creating database "contrib_regression" ============== CREATE DATABASE ALTER DATABASE ============== running regression test queries ============== test test-init-extensions ... ok test test-parseaddress ... ok test test-standardize_address_1 ... ok test test-standardize_address_2 ... ok ===================== All 4 tests passed. =====================
For tiger geocoder, make sure you have postgis and fuzzystrmatch extensions available in your PostgreSQL instance. The address_standardizer tests will also kick in if you built postgis with address_standardizer support:
cd extensions/postgis_tiger_geocoder make install make installcheck
output should look like:
============== dropping database "contrib_regression" ============== DROP DATABASE ============== creating database "contrib_regression" ============== CREATE DATABASE ALTER DATABASE ============== installing fuzzystrmatch ============== CREATE EXTENSION ============== installing postgis ============== CREATE EXTENSION ============== installing postgis_tiger_geocoder ============== CREATE EXTENSION ============== installing address_standardizer ============== CREATE EXTENSION ============== running regression test queries ============== test test-normalize_address ... ok test test-pagc_normalize_address ... ok ===================== All 2 tests passed. =====================
Per installare PostGIS, digitate
make install
Questo comando copierà i file per l'installazione di PostGIS nelle cartelle appropriate, specificate tramite il parametro di configurazione --prefix. In particolare:
I file binari del loader e del dumper vengono installati in [prefix]/bin
.
I file SQL, quali postgis.sql
, vengono installati in [prefix]/share/contrib
.
Le librerie PostGIS vengono installate in [prefix]/lib
.
Se avete eseguito il comando make comments per generare i file postgis_comments.sql
e raster_comments.sql
, installate i file SQL eseguendo
make comments-install
|
Il primo passo per creare un database PostGIS consiste nel creare un semplice database PostgreSQL.
createdb [il_tuo_database]
Molte delle funzioni PostGIS sono scritte nel linguaggio procedurale PL/pgSQL. Pertanto, il prossimo passo per creare un database PostGIS consiste nell'abilitare il linguaggio PL/pgSQL nel vostro database. Per questo deve essere eseguito il comando riportato sotto. Per PostgreSQL 8.4+, in genere PL/pgSQL è già installato
createlang plpgsql [il_tuo_database]
Poi caricate le definizioni degli oggetti e delle funzioni PostGIS, tramite il file di definizioni postgis.sql
(ubicato in [prefix]/share/contrib
come specificato in fase di configurazione).
psql -d [il_tuo_database] -f postgis.sql
Per un insieme completo dei sistemi di coordinate con gli identificatori EPSG potete caricare il file di definizioni spatial_ref_sys.sql
e caricare i valori nella tabella spatial_ref_sys
. Ciò vi consentirà di eseguire le operazioni ST_Transform() sulle geometrie.
psql -d [il_tuo_database] -f spatial_ref_sys.sql
Se volete aggiungere i commenti alle funzioni PostGIS, l'ultimo passo consiste nel caricare postgis_comments.sql
nel vostro database spaziale. I commenti possono essere visti digitando \dd [nome funzione] da una finestra psql.
psql -d [[il_tuo_database] -f postgis_comments.sql
Installazione del supporto raster
psql -d [il_tuo_database] -f rtpostgis.sql
Installazione dei commenti per il supporto raster. Questo fornisce una guida sintetica per ciascuna funzione raster utilizzando psql o PgAdmin o qualsiasi altro strumento di PostgreSQL che possa mostrare i commenti alle funzioni
psql -d [il_tuo_database] -f raster_comments.sql
Installazione del supporto per la topologia
psql -d [il_tuo_database] -f topology/topology.sql
Installa i commenti per il supporto topologico. Questa opzione fornisce una guida rapida per ciascuna delle funzioni e i tipi topologici, tramite psql o PgAdmin o qualsiasi altro strumento PostgreSQL con la capacità di mostrare i commenti delle funzioni
psql -d [il_tuo_database] -f topology/topology_comments.sql
Se prevedete di ripristinare un vecchio backup, da una versione precedente, nel nuovo database, eseguite:
psql -d [il_tuo_database] -f legacy.sql
E' inoltre disponibile un file alternativo ( |
Potete successivamente eseguire uninstall_legacy.sql
per sbarazzarvi delle funzioni obsolete, dopo che avete concluso il ripristino e la pulizia del database.
Se state utilizzando PostgreSQL9.1 o superiore e avete compilato e installato i moduli delle estensioni postgis, potete creare un database spaziale col nuovo sistema.
createdb [il_tuo_database]
The core postgis extension installs PostGIS geometry, geography, raster, spatial_ref_sys and all the functions and comments with a simple:
CREATE EXTENSION postgis;
command.
psql -d [il_tuo_database] -c "CREATE EXTENSION postgis;"
La topologia è compresa in un'estensione separata e può essere installata con il comando:
psql -d [il_tuo_database] -c "CREATE EXTENSION postgis_topology;"
Se prevedete di ripristinare un vecchio backup, da una versione precedente, nel nuovo database, eseguite:
psql -d [il_tuo_database] -f legacy.sql
Potete successivamente eseguire uninstall_legacy.sql
per sbarazzarvi delle funzioni obsolete, dopo che avete concluso il ripristino e la pulizia del database.
The address_standardizer
extension used to be a separate package that required separate download. From PostGIS 2.2 on, it is now bundled in. For more information about the address_standardize, what it does, and how to configure it for your needs, refer to Chapter 12, Address Standardizer.
This standardizer can be used in conjunction with the PostGIS packaged tiger geocoder extension as a replacement for the Normalize_Address discussed. To use as replacement refer to Section 2.8.3, “Using Address Standardizer Extension with Tiger geocoder”. You can also use it as a building block for your own geocoder or use it to standardize your addresses for easier compare of addresses.
The address standardizer relies on PCRE which is usually already installed on many Nix systems, but you can download the latest at: http://www.pcre.org. If during Section 2.4.1, “Configurazione”, PCRE is found, then the address standardizer extension will automatically be built. If you have a custom pcre install you want to use instead, pass to configure --with-pcredir=/path/to/pcre
where /path/to/pcre
is the root folder for your pcre include and lib directories.
For Windows users, the PostGIS 2.1+ bundle is packaged with the address_standardizer already so no need to compile and can move straight to CREATE EXTENSION
step.
Once you have installed, you can connect to your database and run the SQL:
CREATE EXTENSION address_standardizer;
The following test requires no rules, gaz, or lex tables
SELECT num, street, city, state, zip FROM parse_address('1 Devonshire Place, Boston, MA 02109');
Output should be
num | street | city | state | zip -----+------------------------+--------+-------+------- 1 | Devonshire Place PH301 | Boston | MA | 02109
Perl Regex:Assemble is no longer needed for compiling address_standardizer extension since the files it generates are part of the source tree. However if you need to edit the usps-st-city-orig.txt
or usps-st-city-orig.txt usps-st-city-adds.tx
, you need to rebuild parseaddress-stcities.h
which does require Regex:Assemble.
cpan Regexp::Assemble
or if you are on Ubuntu / Debian you might need to do
sudo perl -MCPAN -e "install Regexp::Assemble"
Extras like Tiger geocoder may not be packaged in your PostGIS distribution, but will always be available in the postgis-2.3.0dev.tar.gz file. The instructions provided here are also available in the extras/tiger_geocoder/README
Se siete su Windows e non avete tar installato, potete utilizzare http://www.7-zip.org/ per decomprimere il file tar di PostGIS.
If you are using PostgreSQL 9.1+ and PostGIS 2.1+, you can take advantage of the new extension model for installing tiger geocoder. To do so:
First get binaries for PostGIS 2.1+ or compile and install as usual. This should install the necessary extension files as well for tiger geocoder.
Connect to your database via psql or pgAdmin or some other tool and run the following SQL commands. Note that if you are installing in a database that already has postgis, you don't need to do the first step. If you have fuzzystrmatch
extension already installed, you don't need to do the second step either.
CREATE EXTENSION postgis; CREATE EXTENSION fuzzystrmatch; --this one is optional if you want to use the rules based standardizer (pagc_normalize_address) CREATE EXTENSION address_standardizer; CREATE EXTENSION postgis_tiger_geocoder;
If you already have postgis_tiger_geocoder extension installed, and just want to update to the latest run:
ALTER EXTENSION postgis UPDATE TO "2.3.0dev"; ALTER EXTENSION postgis_topology UPDATE TO "2.3.0dev";
If you made custom entries or changes to tiger.loader_platform
and tiger.loader_variables
you may need to update these.
To confirm your install is working correctly, run this sql in your database:
SELECT na.address, na.streetname,na.streettypeabbrev, na.zip FROM normalize_address('1 Devonshire Place, Boston, MA 02109') AS na;
Which should output
address | streetname | streettypeabbrev | zip ---------+------------+------------------+------- 1 | Devonshire | Pl | 02109
Create a new record in tiger.loader_platform
table with the paths of your executables and server.
So for example to create a profile called debbie that follows sh
convention. You would do:
INSERT INTO tiger.loader_platform(os, declare_sect, pgbin, wget, unzip_command, psql, path_sep, loader, environ_set_command, county_process_command) SELECT 'debbie', declare_sect, pgbin, wget, unzip_command, psql, path_sep, loader, environ_set_command, county_process_command FROM tiger.loader_platform WHERE os = 'sh';
And then edit the paths in the declare_sect column to those that fit Debbie's pg, unzip,shp2pgsql, psql, etc path locations.
If you don't edit this loader_platform
table, it will just contain common case locations of items and you'll have to edit the generated script after the script is generated.
Create a folder called gisdata
on root of server or your local pc if you have a fast network connection to the server. This folder is where the tiger files will be downloaded to and processed. If you are not happy with having the folder on the root of the server, or simply want to change to a different folder for staging, then edit the field staging_fold
in the tiger.loader_variables
table.
Create a folder called temp in the gisdata
folder or whereever you designated the staging_fold
to be. This will be the folder where the loader extracts the downloaded tiger data.
Then run the Loader_Generate_Nation_Script and Loader_Generate_Script SQL functions make sure to use the name of your custom profile and copy the scripts to a .sh or .bat file. So for example to do the nation load and one state using our new profile we would:
SELECT Loader_Generate_Nation_Script('debbie');
SELECT Loader_Generate_Script(ARRAY['MA'], 'debbie');
Eseguite i comandi SQL DROP che vengono generati.
After you are done loading all data or at a stopping point, it's a good idea to analyze all the tiger tables to update the stats (include inherited stats)
SELECT install_missing_indexes(); vacuum analyze verbose tiger.addr; vacuum analyze verbose tiger.edges; vacuum analyze verbose tiger.faces; vacuum analyze verbose tiger.featnames; vacuum analyze verbose tiger.place; vacuum analyze verbose tiger.cousub; vacuum analyze verbose tiger.county; vacuum analyze verbose tiger.state;
If you installed the tiger geocoder without using the extension model, you can convert to the extension model as follows:
Follow instructions in Section 2.8.5, “Aggiornamento del installazione del Geocoder Tiger” for the non-extension model upgrade.
Connect to your database with psql or pgAdmin and run the following command:
CREATE EXTENSION postgis_tiger_geocoder FROM unpackaged;
Anzitutto installate PostGIS utilizzando le istruzioni riportate sopra.
If you don't have an extras folder, download http://postgis.net/stuff/postgis-2.3.0dev.tar.gz
tar xvfz postgis-2.3.0dev.tar.gz
cd postgis-2.3.0dev/extras/tiger_geocoder
Edit the tiger_loader_2015.sql
(or latest loader file you find, unless you want to load different year) to the paths of your executables server etc or alternatively you can update the loader_platform
table once installed. If you don't edit this file or the loader_platform
table, it will just contain common case locations of items and you'll have to edit the generated script after the fact when you run the Loader_Generate_Nation_Script and Loader_Generate_Script SQL functions.
If you are installing Tiger geocoder for the first time edit either the create_geocode.bat
script If you are on windows or the create_geocode.sh
if you are on Linux/Unix/Mac OSX with your PostgreSQL specific settings and run the corresponding script from the commandline.
Verify that you now have a tiger
schema in your database and that it is part of your database search_path. If it is not, add it with a command something along the line of:
ALTER DATABASE geocoder SET search_path=public, tiger;
The normalizing address functionality works more or less without any data except for tricky addresses. Run this test and verify things look like this:
SELECT pprint_addy(normalize_address('202 East Fremont Street, Las Vegas, Nevada 89101')) As pretty_address; pretty_address --------------------------------------- 202 E Fremont St, Las Vegas, NV 89101
One of the many complaints of folks is the address normalizer function Normalize_Address function that normalizes an address for prepping before geocoding. The normalizer is far from perfect and trying to patch its imperfectness takes a vast amount of resources. As such we have integrated with another project that has a much better address standardizer engine. To use this new address_standardizer, you compile the extension as described in Section 2.7, “Installing and Using the address standardizer” and install as an extension in your database.
Once you install this extension in the same database as you have installed postgis_tiger_geocoder
, then the Pagc_Normalize_Address can be used instead of Normalize_Address. This extension is tiger agnostic, so can be used with other data sources such as international addresses. The tiger geocoder extension does come packaged with its own custom versions of rules table ( tiger.pagc_rules
) , gaz table (tiger.pagc_gaz
), and lex table (tiger.pagc_lex
). These you can add and update to improve your standardizing experience for your own needs.
Le istruzioni dettagliate per il caricamento dei dati sono disponibili nel file extras/tiger_geocoder/tiger_2011/README
. Di seguito si illustrano i passaggi generali.
Il processo di caricamento scarica i dati per ciascuno stato richiesto dal sito del censimento, estrare i file, e li carica nelle rispettive tabelle. La tabella di ogni stato eredita le proprietà dalle tabelle definite nello schema tiger
, per cui è sufficiente eseguire le interrogazioni su queste tabelle per accedere a tutti i dati. E' altresì possibile cancellare le tabelle per un dato stato tramite lo script Drop_State_Tables_Generate_Script, qualora aveste necessità di ricaricare uno stato o se questo non vi serve più.
Per caricare i dati avrete bisogno dei seguenti strumenti:
Uno strumento per decomprimere i file zip dal sito web dei censimenti.
Per sistemi tipo Unix: l'eseguibile unzip
è di soiito già installato.
Per Windows, 7-zip è uno strumento libero di compressione/decompressione, che può essere scaricato dal sito http://www.7-zip.org/
Il programam da riga di comando shp2pgsql
, che viene installato di base quando nistallate PostGIS.
wget
è uno strumento per la copia di file da web, solitamente installato nella maggior parte dei sistemi Unix/Linux.
Se siete su Windows, potete reperire i file binari precompilati da http://gnuwin32.sourceforge.net/packages/wget.htm
If you are upgrading from tiger_2010, you'll need to first generate and run Drop_Nation_Tables_Generate_Script. Before you load any state data, you need to load the nation wide data which you do with Loader_Generate_Nation_Script. Which will generate a loader script for you. Loader_Generate_Nation_Script is a one-time step that should be done for upgrading (from 2010) and for new installs.
To load state data refer to Loader_Generate_Script to generate a data load script for your platform for the states you desire. Note that you can install these piecemeal. You don't have to load all the states you want all at once. You can load them as you need them.
Dopo aver caricato i dati per gli stati cui siete interessati, assicuratevi di eseguire:
SELECT install_missing_indexes();
come descritto in Install_Missing_Indexes.
Per vedere se le cose funzionano come dovrebbero, provate a eseguire la geocodifica di un indirizzo nel vostro stato utilizzando Geocode
If you have Tiger Geocoder packaged with 2.0+ already installed, you can upgrade the functions at any time even from an interim tar ball if there are fixes you badly need. This will only work for Tiger geocoder not installed with extensions.
If you don't have an extras folder, download http://postgis.net/stuff/postgis-2.3.0dev.tar.gz
tar xvfz postgis-2.3.0dev.tar.gz
cd postgis-2.3.0dev/extras/tiger_geocoder/tiger_2011
Locate the upgrade_geocoder.bat
script If you are on windows or the upgrade_geocoder.sh
if you are on Linux/Unix/Mac OSX. Edit the file to have your postgis database credentials.
If you are upgrading from 2010 or 2011, make sure to unremark out the loader script line so you get the latest script for loading 2012 data.
Then run th corresponding script from the commandline.
Poi cancellate tutte le tabelle delle nazioni e caricate quelle nuove. Potete generare uno script per la cancellazione con il codice SQL illustrato in Drop_Nation_Tables_Generate_Script
SELECT drop_nation_tables_generate_script();
Eseguite i comandi SQL DROP che vengono generati.
Generate uno script per il caricamento di uno stato con l'istruzione SELECT come illustrato in Loader_Generate_Nation_Script
Per Windows
SELECT loader_generate_nation_script('windows');
Per unix/linux
SELECT loader_generate_nation_script('sh');
Fate riferimento a Section 2.8.4, “Caricamento dei dati Tiger” per istruzioni su come eseguire lo script generato. Questo passaggio deve essere eseguito solo una volta.
Potete avere una combinazione di tabelle degli stati sia per il 2010 che per il 2011 e potete aggiornare ogni stato separatamente. Prima di aggiornare uno stato al 2011, dovete prima eliminare la corrispondente tabella 2012 utilizzando Drop_State_Tables_Generate_Script. |
Alcune distribuzioni pacchettizzate di PostGIS (in particolare i file di installazione Win32 per PostGIS
> 1.1.5) caricano le funzioni PostGIS in un database di esempio denominato template_postgis
. Se il database template_postgis
esiste nella vostra installazione PostgreSQL, allora è possibile per utenti e applicazioni creare database con funzioni spaziali utilizzando un singolo comando. Da notare che in entrambi i casi all'utente del database deve essere riconosciuto il ruolo di creare nuovi database.
Dalla riga di comando:
# createdb -T template_postgis my_spatial_db
Da SQL:
postgres=# CREATE DATABASE my_spatial_db TEMPLATE=template_postgis
L'aggiornamento di database spaziali preesistenti può essere complicato, perché richiede la sostituzione o l'introduzione di nuove definizioni di oggetti PostGIS.
Purtroppo non tutte le definizioni possono essere sostituite facilmente in un database attivo, per cui in alcuni casi la cosa migliore è seguire un processo di scaricamento e ricaricamento dei dati.
PostGIS provides a SOFT UPGRADE procedure for minor or bugfix releases, and a HARD UPGRADE procedure for major releases.
Before attempting to upgrade PostGIS, it is always worth to backup your data. If you use the -Fc flag to pg_dump you will always be able to restore the dump with a HARD UPGRADE.
Se avete installato il database con le estensioni, dovrete eseguire anche l'aggiornamento con il modello delle estensioni. Se avete invece eseguito l'installazione con il vecchio sistema dello script SQL, anche l'aggiornamento dovrà essere eseguito con lo script SQL. Fate riferimento al paragrafo corrispondente.
Questo paragrafo vale solo per coloro che hanno installato PostGIS senza estensioni. Se avete le estensioni e provate a eseguire l'aggiornamento con questo sistema, otterrete messaggi tipo:
non posso eseguire il DROP, dato che c'è una dipendenza dell'estensione PostGIS
After compiling and installing (make install) you should find a postgis_upgrade.sql
and rtpostgis_upgrade.sql
in the installation folders. For example /usr/share/postgresql/9.3/contrib/postgis_upgrade.sql
. Install the postgis_upgrade.sql
. If you have raster functionality installed, you will also need to install the /usr/share/postgresql/9.3/contrib/postgis_upgrade.sql
. If you are moving from PostGIS 1.* to PostGIS 2.* or from PostGIS 2.* prior to r7409, you need to do a HARD UPGRADE.
psql -f postgis_upgrade.sql -d your_spatial_database
La stessa procedura si applica alle estensioni per i raster e per la topologia, i cui file si chiamano rispettivamente rtpostgis_upgrade*.sql
e topology_upgrade*.sql
respectively. Se vi servono:
psql -f rtpostgis_upgrade.sql -d your_spatial_database
psql -f topology_upgrade.sql -d your_spatial_database
If you can't find the |
La funzione PostGIS_Full_Version dovrebbe informarvi della necessita di eseguire questo tipo di aggiornamento attraverso un messaggio tipo "procs need upgrade".
Se inizialmente avete installato PostGIS con le estensioni, anche l'aggiornamento dovrà essere eseguito con le estensioni. Eseguire un aggiornamento di una versione minore con le estensioni è un'operazione abbastanza tranquilla.
ALTER EXTENSION postgis UPDATE TO "2.3.0dev"; ALTER EXTENSION postgis_topology UPDATE TO "2.3.0dev";
Se ottenete un errore tipo:
No migration path defined for ... to 2.3.0dev
Then you'll need to backup your database, create a fresh one as described in Section 2.6, “Creare un database spaziale con le ESTENSIONI” and then restore your backup ontop of this new database.
If you get a notice message like:
Version "2.3.0dev" of extension "postgis" is already installed
Then everything is already up to date and you can safely ignore it. UNLESS you're attempting to upgrade from an SVN version to the next (which doesn't get a new version number); in that case you can append "next" to the version string, and next time you'll need to drop the "next" suffix again:
ALTER EXTENSION postgis UPDATE TO "2.3.0devnext"; ALTER EXTENSION postgis_topology UPDATE TO "2.3.0devnext";
If you installed PostGIS originally without a version specified, you can often skip the reinstallation of postgis extension before restoring since the backup just has |
By HARD UPGRADE we mean full dump/reload of postgis-enabled databases. You need a HARD UPGRADE when PostGIS objects' internal storage changes or when SOFT UPGRADE is not possible. The Release Notes appendix reports for each version whether you need a dump/reload (HARD UPGRADE) to upgrade.
Il processo di scaricamento e ricaricamento dei dati viene assistito dallo script postgis_restore.pl. Questo si fa carico di saltare nello scaricamento tutte le definizioni che fanno parte di PostGIS (comprese quelle obsolete), consentendo così di ripristinare gli schemi e i dati in un database con PostGIS senza ottenere errori legati alla presenza di simboli duplicati e senza portarsi dietro oggetti obsoleti.
Ulteriori istruzioni per gli utenti Windows sono disponibili alla pagina Windows Hard upgrade.
La procedura è la seguente:
Create un dump "con formato personalizzato" del database da aggiornare (che chiameremo olddb
), includendo i blob binari (-b) e con uscita dettagliata (-v). L'utente può essere il proprietario del database, non serve avere privilegi da amministratore.
pg_dump -h localhost -p 5432 -U postgres -Fc -b -v -f "/somepath/olddb.backup" olddb
Eseguire un'installazione da zero di PostGIS in un nuovo database -- chiameremo questo database newdb
. Vi rimandiamo a Section 2.5, “Creare un database spaziale su PostgreSQL con versione inferiore a 9.1” e Section 2.6, “Creare un database spaziale con le ESTENSIONI” per le relative istruzioni.
Le voci spatial_ref_sys presenti nel vostro dump saranno ripristinate, ma non avranno la precedenza su quelle esistenti in spatial_ref_sys. Ciò è per assicurarsi che l'elenco ufficiale venga correttamente trasmesso ai database ripristinati. Se per una qualsiasi ragione volete davvero che le vostre definizioni siano considerate al posto di quelle standard, potete evitare di caricare spatial_ref_sys.sql nel momento in cui create il nuovo database.
Se il vostro database è molto vecchio, o sapete che avete utilizzato nelle vostre viste e nelle vostre funzioni delle funzioni obsolete, potreste dover caricare il file legacy.sql
. Eseguite questa operazione solo se vi serve _veramente_. Considerate la possibilità di aggiornare le vostre viste e le voste funzioni prima di eseguire il dump, se possibile. Le funzioni obsolete possono poi essere rimosse caricando il file uninstall_legacy.sql
.
Ripristinate il vostro backup nel nuovo database newdb
tramite postgis_restore.pl. Eventuali errori imprevisti, se presenti, saranno trascritti da psql sull'uscita standard error. Tenete una traccia di tali errori.
perl utils/postgis_restore.pl "/somepath/olddb.backup" | psql -h localhost -p 5432 -U postgres newdb 2 > errors.txt
Si potrebbero riscontrare errori nei casi seguenti:
Alcune delle vostre viste o funzioni fanno uso di oggetti PostGIS obsoleti. Per risolvere questo problema potete provare a caricare lo script legacy.sql
prima del ripristino, oppure dovrete eseguire il ripristino in una versione di PostGIS che ancora contenga quegli oggetti, e poi ritentare la migrazione dopo aver adattato il codice alla nuova versione. Se il sistema legacy.sql
vi funziona, non dimenticate di sistemare il vostro codice, eliminando i riferimenti alle funzioni obsolete, e poi disinstallate tutti i vecchi riferimenti tramite uninstall_legacy.sql
.
Alcuni record personalizzati di spatial_ref_sys nel file dump hanno un valore non valido per SRID. I valori validi di SRID sono maggiori di 0 e minori di 999000. I valori tra 999000 e 999999 sono riservati ad uso interno, mentre i valori superiori a 999999 non possono essere usati. Tutti i vostri record personalizzati con valori SRID non validi saranno conservati, con quelli >999999 spostati nell'intervallo riservato, ma la tabella spatial_ref_sys perderebbe un controllo e possibilmente anche la propria chiave primaria (quando più SRID non validi vengono convertiti allo stesso valore di SRID riservato).
Per sistemare questo problema dovete copiare il vostro SRS personalizzato in un SRID con un valore valido (nell'intervallo 910000..910999), convertire tutte le tabelle nel nuovo SRID (vedi UpdateGeometrySRID), cancellare la voce non valida da spatial_ref_sys e ricostruire il controllo o in controlli con:
ALTER TABLE spatial_ref_sys ADD CONSTRAINT spatial_ref_sys_srid_check check (srid > 0 AND srid < 999000 );
ALTER TABLE spatial_ref_sys ADD PRIMARY KEY(srid));
Quando l'installazione o l'aggiornamento non vanno come previsto, diverse cose vanno controllate.
Controllate di aver installato PostgreSQL 9.1 o più recente, e che state compilando con il sorgente PostgreSQL nella versione corrispondente alla versione di PostgreSQL che sta girando. Si possono verificare casi di confusione quando la vostra distribuzione (Linux) ha PostgreSQL già installao, o quando avete installato PostgreSQL in precedenza e ve ne siete dimenticati. PostGIS funzionerà solo con PostgreSQL 9.1 o più recente, e si potrebbero ricevere messaggi di errore strani o inattesi se utilizzate una versione più vecchia. Per verificare quale versione di PostgreSQL sta girando, collegatevi al database utilizzando psql ed eseguite la seguente query:
SELECT version();
Se state utilizzando una distribuzione basata su RPM, potete verificare la presenza di pacchetti preinstallati utilizzando il comando rpm con la seguente sintassi: rpm -qa | grep postgresql
Se l'aggiornamento non funziona, assicuratevi di eseguire il ripristino in un database che abbia già PostGIS installato.
SELECT postgis_full_version();
Inoltre, verificate che configure abbia correttamente individuato il percorso e la versione di PostgreSQL, della libreria Proj4 e della libreria GEOS.
L'uscita da configure viene utilizzata per generare il file postgis_config.h
. Controllate che le variabili POSTGIS_PGSQL_VERSION
, POSTGIS_PROJ_VERSION
e POSTGIS_GEOS_VERSION
siano assegnate correttamente.
Il loader e il dumper dei dati sono compilati e installati automaticamente con PostGIS. Per compilarli e installarli manualmente:
# cd postgis-2.3.0dev/loader # make # make install
Il loader si chiama shp2pgsql
e converte gli Shapefile ESRI in formato SQL adatto al caricamento in PostGIS/PostgreSQL. Il dumper si chiama pgsql2shp
e converte le tabelle (o le query) di PostGIS in formato Shapefile ESRI. Per una documentazione più estesa, si rimanda alla guida in linea e alle pagine del manuale.
3.1. | Where can I find tutorials, guides and workshops on working with PostGIS | |||
OpenGeo has a step by step tutorial guide workshop Introduction to PostGIS. It includes packaged data as well as intro to working with OpenGeo Suite. It is probably the best tutorial on PostGIS. BostonGIS also has a PostGIS almost idiot's guide on getting started. That one is more focused on the windows user. | ||||
3.2. | Le mie applicazioni e i miei strumenti desktop funzionavano con PostGIS 1.5, ma non funzionano con PostGIS 2.0. Come posso risolvere la cosa? | |||
Diverse funzioni obsolete sono state eliminate dal codice PostGIS nella versione 2.0. Ciò ha avuto un effetto sulle applicazioni, oltre che su strumenti di terze parti quali Geoserver, MapServer, QuantumGIS e OpenJump, per citarne alcuni. Ci sono un paio di modi per risolvere il problema. Per le applicazioni di terze parti, potere provare ad aggiornarle alla versione più recente, in cui molti di questi problemi sono stati sistemati. Per il codice da che hai sviluppato tu, puoi modificare il codice in modo da non utilizzare le funzioni eliminate. [TO BE COMPLETED HAVING CLARIFIED ISSUE] Il file | ||||
3.3. | When I load OpenStreetMap data with osm2pgsql, I'm getting an error failed: ERROR: operator class "gist_geometry_ops" does not exist for access method "gist" Error occurred. This worked fine in PostGIS 1.5. | |||
In PostGIS 2, the default geometry operator class gist_geometry_ops was changed to gist_geometry_ops_2d and the gist_geometry_ops was completely removed. This was done because PostGIS 2 also introduced Nd spatial indexes for 3D support and the old name was deemed confusing and a misnomer. Some older applications that as part of the process create tables and indexes, explicitly referenced the operator class name. This was unnecessary if you want the default 2D index. So if you manage said good, change index creation from: BAD: CREATE INDEX idx_my_table_geom ON my_table USING gist(geom gist_geometry_ops); To GOOD: CREATE INDEX idx_my_table_geom ON my_table USING gist(geom); The only case where you WILL need to specify the operator class is if you want a 3D spatial index as follows: CREATE INDEX idx_my_super3d_geom ON my_super3d USING gist(geom gist_geometry_ops_nd); If you are unfortunate to be stuck with compiled code you can't change that has the old gist_geometry_ops hard-coded, then you can create the old class using the | ||||
3.4. | Sto usando PostgreSQL 9.0 e non posso più leggere le geometrie in OpenJump, Safe FME e altri strumenti. Come mai? | |||
In PostgreSQL 9.0+, la codifica di default per i dati di tipo bytea è stata modificata in esadecimale e i vecchi driver JDBC considerano sempre il formato di escape. Ciò ha avuto un effetto su alcune applicazioni, quali le applicazioni Java con i driver JDBC più vecchi o le applicazioni .NET che utilizzano il vecchio driver npgsql, che si aspetta di incontrare il vecchio comportamento di ST_AsBinary. Ci sono due approcci per far funzionare nuovamente il tutto. Puoi aggiornare il driver JDBC alla versione PostgreSQL 9.0 più recente dal sito http://jdbc.postgresql.org/download.html Se stai usando un'applicazione .NET, puoi utilizzare Npgsql 2.0.11 o versioni superiori, scaricandolo da http://pgfoundry.org/frs/?group_id=1000140 e seguendo le istruzioni descritte nel blog di Francisco Figueiredo su NpgSQL 2.0.11 Se non è possibile aggiornare il tuo driver PostgreSQL, puoi ripristinare il comportamento precedente con la seguente modifica: ALTER DATABASE mypostgisdb SET bytea_output='escape'; | ||||
3.5. | Ho provato a usare PgAdmin per visualizzare la mia colonna della geometria e questa risulta vuota. Perché? | |||
PgAdmin non mostra nulla nel caso di grandi geometrie. I modi migliori per verificare che vi siano dati presenti nelle vostre colonne di geometria sono: -- se tutti i tuoi campi geometrici sono popolati, questa query non dovrebbe restituire alcun record SELECT somefield FROM mytable WHERE geom IS NULL; -- Per conoscere giusto la dimensione di una geometria, è possibile eseguire una query in una forma --che vi dirà il più elevato numero di punti in una colonna di geometria SELECT MAX(ST_NPoints(geom)) FROM sometable; | ||||
3.6. | Quale tipo di oggetti geometrici posso memorizzare? | |||
You can store Point, LineString, Polygon, MultiPoint, MultiLineString, MultiPolygon, and GeometryCollection geometries. In PostGIS 2.0 and above you can also store TINS and Polyhedral Surfaces in the basic geometry type. These are specified in the Open GIS Well Known Text Format (with Z, M, and ZM extensions). There are three data types currently supported. The standard OGC geometry data type which uses a planar coordinate system for measurement, the geography data type which uses a geodetic coordinate system, with calculations on either a sphere or spheroid. The newest family member of the PostGIS spatial type family is raster for storing and analyzing raster data. Raster has its very own FAQ. Refer to Chapter 10, Domande frequenti su PostGIS raster and Chapter 9, Raster Reference for more details. | ||||
3.7. | Sono confuso. Quale tipo di dati dovrei usare, quello geometrico o quello geografico? | |||
Short Answer: geography is a newer data type that supports long range distances measurements, but most computations on it are slower than they are on geometry. If you use geography, you don't need to learn much about planar coordinate systems. Geography is generally best if all you care about is measuring distances and lengths and you have data from all over the world. Geometry data type is an older data type that has many more functions supporting it, enjoys greater support from third party tools, and operations on it are generally faster -- sometimes as much as 10 fold faster for larger geometries. Geometry is best if you are pretty comfortable with spatial reference systems or you are dealing with localized data where all your data fits in a single spatial reference system (SRID), or you need to do a lot of spatial processing. Note: It is fairly easy to do one-off conversions between the two types to gain the benefits of each. Refer to Section 14.10, “PostGIS Function Support Matrix” to see what is currently supported and what is not. Risposta lunga: ti rimandiamo alla discussione più estesa Section 4.2.2, “When to use Geography Data type over Geometry data type” e matrice dei tipi di funzione. | ||||
3.8. | Ho altre domande più profonde sulla geografia, ad esempio qual è la massima dimensione di una regione geografica che posso inserire in una colonna di tipo geography per continuare ad avere risultati ragionevoli. Ci sono limitazioni ai poli, tutto deve essere in un unico emisfero (come per SQL Server 2008)? | |||
Le tue domande sono troppo profonde e complesse per avere una risposta in questa sezione del documento. Ti rimandiamo alle nostre Section 4.2.3, “Geography Advanced FAQ”. | ||||
3.9. | Come posso inserire un oggetto GIS nel database? | |||
Anzitutto devi creare una tabella con una colonna di tipo "geometry" o "geography" per contenere i dati GIS. Memorizzare i dati di tipo geografico e un po' diverso dal memorizzare dati geometrici. Ti rimandiamo a Section 4.2.1, “Geography Basics” per i dettagli sulla gestione dei dati geografici. Per la geometria: collegati al database con CREATE TABLE gtest ( gid serial primary key, name varchar(20) , geom geometry(LINESTRING) ); If the geometry column definition fails, you probably have not loaded the PostGIS functions and objects into this database or are using a pre-2.0 version of PostGIS. See the Section 2.4, “Compiling and Install from Source: Detailed”. Fatto questo, puoi inserire una geometria nella tabella utilizzando un comando di INSERT SQL. L'oggetto GIS vero e proprio è formattato utilizzando le specifiche "well-known text" dell'OpenGIS Consortium: INSERT INTO gtest (ID, NAME, GEOM) VALUES ( 1, 'First Geometry', ST_GeomFromText('LINESTRING(2 3,4 5,6 5,7 8)') ); Per ulteriori informazioni su altri oggetti GIS, si rimanda alla guida di riferimento sugli oggetti. Per visualizzare i dati GIS nella tabella: SELECT id, name, ST_AsText(geom) AS geom FROM gtest; Il valore restituito dovrebbe essere qualcosa tipo: id | name | geom ----+----------------+----------------------------- 1 | First Geometry | LINESTRING(2 3,4 5,6 5,7 8) (1 row) | ||||
3.10. | Come posso creare una query spaziale? | |||
Nello stesso modo in cui creeresti qualsiasi altra query per il database, come combinazione SQL di valori, funzioni e test booleani. Per le query spaziali, è bene tenere a mente due aspetti: se esiste un indice spaziale che puoi utilizzare, e se stai eseguendo calcoli impegnativi su un grande numero di geometrie. In generale, vorrai considerare l'uso dell'operatore di intersezione (&&) che verifica se le bounding box degli elementi si intersecano. La ragione per cui l'operatore &&) è utile è perché questo operatore ne farà uso, accelerando di molto le query. Utilizzerai anche funzioni spaziali quali Distance(), ST_Intersects(), ST_Contains() and ST_Within(), tra le altre, per ridurre l'insieme dei risultati della ricerca. Molte interrogazioni spaziali comprendono sia un test legato a un indice, che un test spaziale. Il test legato all'indice serve a limitare il numero di tuple a quelle che potrebbero rispondere alle condizioni richieste. Le funzioni spaziali vanno poi a verificare in modo più esatto le condizioni. SELECT id, the_geom FROM thetable WHERE ST_Contains(the_geom,'POLYGON((0 0, 0 10, 10 10, 10 0, 0 0))'); | ||||
3.11. | Come posso accelerare le query spaziali su grandi tabelle? | |||
La velocità delle query su tabelle di grandi dimensioni è la raison d'etre dei database spaziali (assieme al supporto delle transazioni), per cui avere un buon indice è importante. Per creare un indice spaziale per una tabella con colonna CREATE INDEX [indexname] ON [tablename] USING GIST ( [geometrycolumn] ); L'opzione "USING GIST" dice al server di utilizzare un indice GiST (Generalized Search Tree).
You should also ensure that the PostgreSQL query planner has enough information about your index to make rational decisions about when to use it. To do this, you have to "gather statistics" on your geometry tables. Per PostgreSQL 8.0.x e versioni superiori, è sufficiente eseguire il comando VACUUM ANALYZE. Per PostgreSQL 7.4.x e versioni precedenti, eseguite il comando SELECT UPDATE_GEOMETRY_STATS(). | ||||
3.12. | Perché gli indici R-Tree di PostgreSQL non sono supportati? | |||
Le prime versioni di PostGIS usavano gli indici R-Tree di PostgreSQL. Tuttavia questi sono stati completamente abbandonati a partire dalla versione 0.6, e l'indicizzazione spaziale viene fornita con un approccio R-Tree attraverso GiST. I nostri test hanno mostrato che la velocità di ricerca per gli indici nativi con R-tree e quelli con GiST sono paragonabili. Gli indici nativi di PostgreSQL hanno due limitazioni che li rendono non desiderabili quando si tratta di elementi GIS (si noti che queste limitazioni sono dovute all'attuale implementazione degli indici R-Tree da parte di PostgreSQL e non al concetto di R-Tree in generale):
| ||||
3.13. | Perché dovrei usare la funzione | |||
Se non volete utilizzare le funzioni di supporto per OpenGIS, non siete obbligati. E' sufficiente creare le tabelle come nelle versioni più vecchie, definendo le colonne per la geometria nel comando CREATE. Tutte le geometrie avranno uno SRID uguale a -1, e le tabelle con i metadati OpenGIS non saranno compilate correttamente. Tuttavia, ciò farà sì che la maggior parte delle applicazioni basate su PostGIS abbia problemi, e in genere è consigliabile usare MapServer è un'applicazione che fa uso dei medatati | ||||
3.14. | Qual è il modo migliore per trovare tutti gli oggetti entro un dato raggio da un altro oggetto? | |||
Per utilizzare il database in modo efficiente, la cosa migliore è eseguire le query con un raggio assieme al test sulla bounding box: il test sulla bounding box ricorre all'indice spaziale, dando un accesso velocizzato a un sottoinsieme dei dati cui poi viene applicato il test sul raggio. La funzione Per esempio, per trovare tutti gli oggetti entro 100 metri dal punto POINT(1000 1000), la seguente query funzionerebbe bene: SELECT * FROM geotable WHERE ST_DWithin(geocolumn, 'POINT(1000 1000)', 100.0); | ||||
3.15. | Come posso eseguire una riproiezione di coordinate come parte di una query? | |||
Per eseguire una riproiezione, sia il sistema di riferimento di coordinate di origine che quello di destinazione devono essere definiti nella tabella SPATIAL_REF_SYS, e le geometrie da riproiettare devono già avere uno SRID assegnato. Fatto questo, la riproiezione è semplice come riferirsi allo SRID desiderato per la destinazione. L'esempio proietta una geometria in NAD 83 long lat. Il caso sotto funzionerà solamente se lo SRID di the_geom non è -1 (il riferimento spaziale non deve essere indefinito). SELECT ST_Transform(the_geom,4269) FROM geotable; | ||||
3.16. | Ho eseguito una ST_AsEWKT e ST_AsText su una geometria piuttosto grossa ed ho ottenuto un campo vuoto. Come mai? | |||
Probabilmente stato usando PgAdmin o qualche altro strumento che non dà in uscita testi di grandi dimensioni. Se la vostra geometria è abbastanza grande, risulterà vuota in questi strumenti. Se dovete vederla o dovete vedere l'uscita in formato WKT potete usare PSQL, --Per verificare il numero di geometrie che sono effettivamente vuote SELECT count(gid) FROM geotable WHERE the_geom IS NULL; | ||||
3.17. | Quando eseguo ST_Intersects, mi dice che le mie due geometrie non si intersecano, mentre SO CHE LO FANNO. Come mai? | |||
Ciò in genere accade in due casi frequenti. O la tua geometria non è valida (controlla ST_IsValid), oppure presumi che si intersechi perche ST_AsText esegue un troncamento dei numeri e hai molti decimali che non vengono mostrati. | ||||
3.18. | Sto rilasciando software che utilizza PostGIS. Significa che dovrà utilizzare per questo la licenza GPL come PostGIS? Dovrò pubblicare tutto il mio codice se uso PostGIS? | |||
Quasi sicuramente no. Per esempio, considera un database Oracle su Linux. Linux è GPL. Oracle non lo è. L'Oracle che gira su Linux deve essere distribuito con licenza GPL? No. Pertanto il tuo software può utilizzare PostgreSQL/PostGIS quanto vuole, ed essere rilasciato sotto qualsiasi tipo di licenza. L'unica eccezione sarebbe se hai eseguito delle modifiche al sorgente di PostGIS, e hai distribuito la versione modificata di PostGIS. In questo caso devi condividere il codice del PostGIS modificato (ma non il codice delle applicazioni che girano su questo). Anche in questo caso particolare, dovrai distribuire il codice sorgente a coloro cui hai distribuito i file binari. La licenza GPL non richiede che tu pubblichi il sorgente, ma solo che tu lo condivida con coloro a cui dai i file binari. |
Table of Contents
The GIS objects supported by PostGIS are a superset of the "Simple Features" defined by the OpenGIS Consortium (OGC). As of version 0.9, PostGIS supports all the objects and functions specified in the OGC "Simple Features for SQL" specification.
PostGIS extends the standard with support for 3DZ,3DM and 4D coordinates.
The OpenGIS specification defines two standard ways of expressing spatial objects: the Well-Known Text (WKT) form and the Well-Known Binary (WKB) form. Both WKT and WKB include information about the type of the object and the coordinates which form the object.
Examples of the text representations (WKT) of the spatial objects of the features are as follows:
POINT(0 0)
LINESTRING(0 0,1 1,1 2)
POLYGON((0 0,4 0,4 4,0 4,0 0),(1 1, 2 1, 2 2, 1 2,1 1))
MULTIPOINT((0 0),(1 2))
MULTILINESTRING((0 0,1 1,1 2),(2 3,3 2,5 4))
MULTIPOLYGON(((0 0,4 0,4 4,0 4,0 0),(1 1,2 1,2 2,1 2,1 1)), ((-1 -1,-1 -2,-2 -2,-2 -1,-1 -1)))
GEOMETRYCOLLECTION(POINT(2 3),LINESTRING(2 3,3 4))
The OpenGIS specification also requires that the internal storage format of spatial objects include a spatial referencing system identifier (SRID). The SRID is required when creating spatial objects for insertion into the database.
Input/Output of these formats are available using the following interfaces:
bytea WKB = ST_AsBinary(geometry); text WKT = ST_AsText(geometry); geometry = ST_GeomFromWKB(bytea WKB, SRID); geometry = ST_GeometryFromText(text WKT, SRID);
For example, a valid insert statement to create and insert an OGC spatial object would be:
INSERT INTO geotable ( the_geom, the_name ) VALUES ( ST_GeomFromText('POINT(-126.4 45.32)', 312), 'A Place');
OGC formats only support 2d geometries, and the associated SRID is *never* embedded in the input/output representations.
PostGIS extended formats are currently superset of OGC one (every valid WKB/WKT is a valid EWKB/EWKT) but this might vary in the future, specifically if OGC comes out with a new format conflicting with our extensions. Thus you SHOULD NOT rely on this feature!
PostGIS EWKB/EWKT add 3dm,3dz,4d coordinates support and embedded SRID information.
Examples of the text representations (EWKT) of the extended spatial objects of the features are as follows. The * ones are new in this version of PostGIS:
POINT(0 0 0) -- XYZ
SRID=32632;POINT(0 0) -- XY with SRID
POINTM(0 0 0) -- XYM
POINT(0 0 0 0) -- XYZM
SRID=4326;MULTIPOINTM(0 0 0,1 2 1) -- XYM with SRID
MULTILINESTRING((0 0 0,1 1 0,1 2 1),(2 3 1,3 2 1,5 4 1))
POLYGON((0 0 0,4 0 0,4 4 0,0 4 0,0 0 0),(1 1 0,2 1 0,2 2 0,1 2 0,1 1 0))
MULTIPOLYGON(((0 0 0,4 0 0,4 4 0,0 4 0,0 0 0),(1 1 0,2 1 0,2 2 0,1 2 0,1 1 0)),((-1 -1 0,-1 -2 0,-2 -2 0,-2 -1 0,-1 -1 0)))
GEOMETRYCOLLECTIONM( POINTM(2 3 9), LINESTRINGM(2 3 4, 3 4 5) )
MULTICURVE( (0 0, 5 5), CIRCULARSTRING(4 0, 4 4, 8 4) )
POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )
TRIANGLE ((0 0, 0 9, 9 0, 0 0))
TIN( ((0 0 0, 0 0 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 0 0 0)) )
Input/Output of these formats are available using the following interfaces:
bytea EWKB = ST_AsEWKB(geometry); text EWKT = ST_AsEWKT(geometry); geometry = ST_GeomFromEWKB(bytea EWKB); geometry = ST_GeomFromEWKT(text EWKT);
For example, a valid insert statement to create and insert a PostGIS spatial object would be:
INSERT INTO geotable ( the_geom, the_name ) VALUES ( ST_GeomFromEWKT('SRID=312;POINTM(-126.4 45.32 15)'), 'A Place' )
The "canonical forms" of a PostgreSQL type are the representations you get with a simple query (without any function call) and the one which is guaranteed to be accepted with a simple insert, update or copy. For the postgis 'geometry' type these are:
- Output - binary: EWKB ascii: HEXEWKB (EWKB in hex form) - Input - binary: EWKB ascii: HEXEWKB|EWKT
For example this statement reads EWKT and returns HEXEWKB in the process of canonical ascii input/output:
=# SELECT 'SRID=4;POINT(0 0)'::geometry; geometry ---------------------------------------------------- 01010000200400000000000000000000000000000000000000 (1 row)
The SQL Multimedia Applications Spatial specification extends the simple features for SQL spec by defining a number of circularly interpolated curves.
The SQL-MM definitions include 3dm, 3dz and 4d coordinates, but do not allow the embedding of SRID information.
The well-known text extensions are not yet fully supported. Examples of some simple curved geometries are shown below:
CIRCULARSTRING(0 0, 1 1, 1 0)
CIRCULARSTRING(0 0, 4 0, 4 4, 0 4, 0 0)
The CIRCULARSTRING is the basic curve type, similar to a LINESTRING in the linear world. A single segment required three points, the start and end points (first and third) and any other point on the arc. The exception to this is for a closed circle, where the start and end points are the same. In this case the second point MUST be the center of the arc, ie the opposite side of the circle. To chain arcs together, the last point of the previous arc becomes the first point of the next arc, just like in LINESTRING. This means that a valid circular string must have an odd number of points greated than 1.
COMPOUNDCURVE(CIRCULARSTRING(0 0, 1 1, 1 0),(1 0, 0 1))
A compound curve is a single, continuous curve that has both curved (circular) segments and linear segments. That means that in addition to having well-formed components, the end point of every component (except the last) must be coincident with the start point of the following component.
CURVEPOLYGON(CIRCULARSTRING(0 0, 4 0, 4 4, 0 4, 0 0),(1 1, 3 3, 3 1, 1 1))
Example compound curve in a curve polygon: CURVEPOLYGON(COMPOUNDCURVE(CIRCULARSTRING(0 0,2 0, 2 1, 2 3, 4 3),(4 3, 4 5, 1 4, 0 0)), CIRCULARSTRING(1.7 1, 1.4 0.4, 1.6 0.4, 1.6 0.5, 1.7 1) )
A CURVEPOLYGON is just like a polygon, with an outer ring and zero or more inner rings. The difference is that a ring can take the form of a circular string, linear string or compound string.
As of PostGIS 1.4 PostGIS supports compound curves in a curve polygon.
MULTICURVE((0 0, 5 5),CIRCULARSTRING(4 0, 4 4, 8 4))
The MULTICURVE is a collection of curves, which can include linear strings, circular strings or compound strings.
MULTISURFACE(CURVEPOLYGON(CIRCULARSTRING(0 0, 4 0, 4 4, 0 4, 0 0),(1 1, 3 3, 3 1, 1 1)),((10 10, 14 12, 11 10, 10 10),(11 11, 11.5 11, 11 11.5, 11 11)))
This is a collection of surfaces, which can be (linear) polygons or curve polygons.
PostGIS prior to 1.4 does not support compound curves in a curve polygon, but PostGIS 1.4 and above do support the use of Compound Curves in a Curve Polygon. |
All floating point comparisons within the SQL-MM implementation are performed to a specified tolerance, currently 1E-8. |
The geography type provides native support for spatial features represented on "geographic" coordinates (sometimes called "geodetic" coordinates, or "lat/lon", or "lon/lat"). Geographic coordinates are spherical coordinates expressed in angular units (degrees).
The basis for the PostGIS geometry type is a plane. The shortest path between two points on the plane is a straight line. That means calculations on geometries (areas, distances, lengths, intersections, etc) can be calculated using cartesian mathematics and straight line vectors.
The basis for the PostGIS geographic type is a sphere. The shortest path between two points on the sphere is a great circle arc. That means that calculations on geographies (areas, distances, lengths, intersections, etc) must be calculated on the sphere, using more complicated mathematics. For more accurate measurements, the calculations must take the actual spheroidal shape of the world into account, and the mathematics becomes very complicated indeed.
Because the underlying mathematics is much more complicated, there are fewer functions defined for the geography type than for the geometry type. Over time, as new algorithms are added, the capabilities of the geography type will expand.
One restriction is that it only supports WGS 84 long lat (SRID:4326). It uses a new data type called geography. None of the GEOS functions support this new type. As a workaround one can convert back and forth between geometry and geography types.
The new geography type uses the PostgreSQL 8.3+ typmod definition format so that a table with a geography field can be added in a single step. All the standard OGC formats except for curves are supported.
The geography type only supports the simplest of simple features. Standard geometry type data will autocast to geography if it is of SRID 4326. You can also use the EWKT and EWKB conventions to insert data.
POINT: Creating a table with 2d point geometry:
CREATE TABLE testgeog(gid serial PRIMARY KEY, the_geog geography(POINT,4326) );
Creating a table with z coordinate point
CREATE TABLE testgeog(gid serial PRIMARY KEY, the_geog geography(POINTZ,4326) );
LINESTRING
POLYGON
MULTIPOINT
MULTILINESTRING
MULTIPOLYGON
GEOMETRYCOLLECTION
The new geography fields don't get registered in the geometry_columns
. They get registered in a new view called geography_columns which is a view against the system catalogs so is always automatically kept up to date without need for an AddGeom... like function.
Now, check the "geography_columns" view and see that your table is listed.
You can create a new table with a GEOGRAPHY column using the CREATE TABLE syntax. Unlike GEOMETRY, there is no need to run a separate AddGeometryColumns() process to register the column in metadata.
CREATE TABLE global_points ( id SERIAL PRIMARY KEY, name VARCHAR(64), location GEOGRAPHY(POINT,4326) );
Note that the location column has type GEOGRAPHY and that geography type supports two optional modifier: a type modifier that restricts the kind of shapes and dimensions allowed in the column; an SRID modifier that restricts the coordinate reference identifier to a particular number.
Allowable values for the type modifier are: POINT, LINESTRING, POLYGON, MULTIPOINT, MULTILINESTRING, MULTIPOLYGON. The modifier also supports dimensionality restrictions through suffixes: Z, M and ZM. So, for example a modifier of 'LINESTRINGM' would only allow line strings with three dimensions in, and would treat the third dimension as a measure. Similarly, 'POINTZM' would expect four dimensional data.
The SRID modifier is currently of limited use: only 4326 (WGS84) is allowed as a value. If you do not specify an SRID, the a value 0 (undefined spheroid) will be used, and all calculations will proceed using WGS84 anyways.
In the future, alternate SRIDs will allow calculations on spheroids other than WGS84.
Once you have created your table, you can see it in the GEOGRAPHY_COLUMNS table:
-- See the contents of the metadata view SELECT * FROM geography_columns;
You can insert data into the table the same as you would if it was using a GEOMETRY column:
-- Add some data into the test table INSERT INTO global_points (name, location) VALUES ('Town', ST_GeographyFromText('SRID=4326;POINT(-110 30)') ); INSERT INTO global_points (name, location) VALUES ('Forest', ST_GeographyFromText('SRID=4326;POINT(-109 29)') ); INSERT INTO global_points (name, location) VALUES ('London', ST_GeographyFromText('SRID=4326;POINT(0 49)') );
Creating an index works the same as GEOMETRY. PostGIS will note that the column type is GEOGRAPHY and create an appropriate sphere-based index instead of the usual planar index used for GEOMETRY.
-- Index the test table with a spherical index CREATE INDEX global_points_gix ON global_points USING GIST ( location );
Query and measurement functions use units of meters. So distance parameters should be expressed in meters, and return values should be expected in meters (or square meters for areas).
-- Show a distance query and note, London is outside the 1000km tolerance SELECT name FROM global_points WHERE ST_DWithin(location, ST_GeographyFromText('SRID=4326;POINT(-110 29)'), 1000000);
You can see the power of GEOGRAPHY in action by calculating the how close a plane flying from Seattle to London (LINESTRING(-122.33 47.606, 0.0 51.5)) comes to Reykjavik (POINT(-21.96 64.15)).
-- Distance calculation using GEOGRAPHY (122.2km) SELECT ST_Distance('LINESTRING(-122.33 47.606, 0.0 51.5)'::geography, 'POINT(-21.96 64.15)':: geography);
-- Distance calculation using GEOMETRY (13.3 "degrees") SELECT ST_Distance('LINESTRING(-122.33 47.606, 0.0 51.5)'::geometry, 'POINT(-21.96 64.15)':: geometry);
The GEOGRAPHY type calculates the true shortest distance over the sphere between Reykjavik and the great circle flight path between Seattle and London.
Great Circle mapper The GEOMETRY type calculates a meaningless cartesian distance between Reykjavik and the straight line path from Seattle to London plotted on a flat map of the world. The nominal units of the result might be called "degrees", but the result doesn't correspond to any true angular difference between the points, so even calling them "degrees" is inaccurate.
The new GEOGRAPHY type allows you to store data in longitude/latitude coordinates, but at a cost: there are fewer functions defined on GEOGRAPHY than there are on GEOMETRY; those functions that are defined take more CPU time to execute.
The type you choose should be conditioned on the expected working area of the application you are building. Will your data span the globe or a large continental area, or is it local to a state, county or municipality?
If your data is contained in a small area, you might find that choosing an appropriate projection and using GEOMETRY is the best solution, in terms of performance and functionality available.
If your data is global or covers a continental region, you may find that GEOGRAPHY allows you to build a system without having to worry about projection details. You store your data in longitude/latitude, and use the functions that have been defined on GEOGRAPHY.
If you don't understand projections, and you don't want to learn about them, and you're prepared to accept the limitations in functionality available in GEOGRAPHY, then it might be easier for you to use GEOGRAPHY than GEOMETRY. Simply load your data up as longitude/latitude and go from there.
Refer to Section 14.10, “PostGIS Function Support Matrix” for compare between what is supported for Geography vs. Geometry. For a brief listing and description of Geography functions, refer to Section 14.3, “PostGIS Geography Support Functions”
The OpenGIS "Simple Features Specification for SQL" defines standard GIS object types, the functions required to manipulate them, and a set of meta-data tables. In order to ensure that meta-data remain consistent, operations such as creating and removing a spatial column are carried out through special procedures defined by OpenGIS.
There are two OpenGIS meta-data tables: SPATIAL_REF_SYS
and GEOMETRY_COLUMNS
. The SPATIAL_REF_SYS
table holds the numeric IDs and textual descriptions of coordinate systems used in the spatial database.
The spatial_ref_sys table is a PostGIS included and OGC compliant database table that lists over 3000 known spatial reference systems and details needed to transform/reproject between them.
Although the PostGIS spatial_ref_sys table contains over 3000 of the more commonly used spatial reference system definitions that can be handled by the proj library, it does not contain all known to man and you can even define your own custom projection if you are familiar with proj4 constructs. Keep in mind that most spatial reference systems are regional and have no meaning when used outside of the bounds they were intended for.
An excellent resource for finding spatial reference systems not defined in the core set is http://spatialreference.org/
Some of the more commonly used spatial reference systems are: 4326 - WGS 84 Long Lat, 4269 - NAD 83 Long Lat, 3395 - WGS 84 World Mercator, 2163 - US National Atlas Equal Area, Spatial reference systems for each NAD 83, WGS 84 UTM zone - UTM zones are one of the most ideal for measurement, but only cover 6-degree regions.
Various US state plane spatial reference systems (meter or feet based) - usually one or 2 exists per US state. Most of the meter ones are in the core set, but many of the feet based ones or ESRI created ones you will need to pull from spatialreference.org.
For details on determining which UTM zone to use for your area of interest, check out the utmzone PostGIS plpgsql helper function.
The SPATIAL_REF_SYS
table definition is as follows:
CREATE TABLE spatial_ref_sys ( srid INTEGER NOT NULL PRIMARY KEY, auth_name VARCHAR(256), auth_srid INTEGER, srtext VARCHAR(2048), proj4text VARCHAR(2048) )
The SPATIAL_REF_SYS
columns are as follows:
An integer value that uniquely identifies the Spatial Referencing System (SRS) within the database.
The name of the standard or standards body that is being cited for this reference system. For example, "EPSG" would be a valid AUTH_NAME
.
The ID of the Spatial Reference System as defined by the Authority cited in the AUTH_NAME
. In the case of EPSG, this is where the EPSG projection code would go.
The Well-Known Text representation of the Spatial Reference System. An example of a WKT SRS representation is:
PROJCS["NAD83 / UTM Zone 10N", GEOGCS["NAD83", DATUM["North_American_Datum_1983", SPHEROID["GRS 1980",6378137,298.257222101] ], PRIMEM["Greenwich",0], UNIT["degree",0.0174532925199433] ], PROJECTION["Transverse_Mercator"], PARAMETER["latitude_of_origin",0], PARAMETER["central_meridian",-123], PARAMETER["scale_factor",0.9996], PARAMETER["false_easting",500000], PARAMETER["false_northing",0], UNIT["metre",1] ]
For a listing of EPSG projection codes and their corresponding WKT representations, see http://www.opengeospatial.org/. For a discussion of WKT in general, see the OpenGIS "Coordinate Transformation Services Implementation Specification" at http://www.opengeospatial.org/standards. For information on the European Petroleum Survey Group (EPSG) and their database of spatial reference systems, see http://www.epsg.org.
PostGIS uses the Proj4 library to provide coordinate transformation capabilities. The PROJ4TEXT
column contains the Proj4 coordinate definition string for a particular SRID. For example:
+proj=utm +zone=10 +ellps=clrk66 +datum=NAD27 +units=m
For more information about, see the Proj4 web site at http://trac.osgeo.org/proj/. The spatial_ref_sys.sql
file contains both SRTEXT
and PROJ4TEXT
definitions for all EPSG projections.
In versions of PostGIS prior to 2.0.0, geometry_columns was a table that could be directly edited, and sometimes got out of synch with the actual definition of the geometry columns. In PostGIS 2.0.0, GEOMETRY_COLUMNS
became a view with the same front-facing structure as prior versions, but reading from database system catalogs Its structure is as follows:
\d geometry_columns
View "public.geometry_columns" Column | Type | Modifiers -------------------+------------------------+----------- f_table_catalog | character varying(256) | f_table_schema | character varying(256) | f_table_name | character varying(256) | f_geometry_column | character varying(256) | coord_dimension | integer | srid | integer | type | character varying(30) |
The column meanings have not changed from prior versions and are:
The fully qualified name of the feature table containing the geometry column. Note that the terms "catalog" and "schema" are Oracle-ish. There is not PostgreSQL analogue of "catalog" so that column is left blank -- for "schema" the PostgreSQL schema name is used (public
is the default).
The name of the geometry column in the feature table.
The spatial dimension (2, 3 or 4 dimensional) of the column.
The ID of the spatial reference system used for the coordinate geometry in this table. It is a foreign key reference to the SPATIAL_REF_SYS
.
The type of the spatial object. To restrict the spatial column to a single type, use one of: POINT, LINESTRING, POLYGON, MULTIPOINT, MULTILINESTRING, MULTIPOLYGON, GEOMETRYCOLLECTION or corresponding XYM versions POINTM, LINESTRINGM, POLYGONM, MULTIPOINTM, MULTILINESTRINGM, MULTIPOLYGONM, GEOMETRYCOLLECTIONM. For heterogeneous (mixed-type) collections, you can use "GEOMETRY" as the type.
This attribute is (probably) not part of the OpenGIS specification, but is required for ensuring type homogeneity. |
Creating a table with spatial data, can be done in one step. As shown in the following example which creates a roads table with a 2D linestring geometry column in WGS84 long lat
CREATE TABLE ROADS ( ID int4 , ROAD_NAME varchar(25), geom geometry(LINESTRING,4326) );
We can add additional columns using standard ALTER TABLE command as we do in this next example where we add a 3-D linestring.
ALTER TABLE roads ADD COLUMN geom2 geometry(LINESTRINGZ,4326);
For backwards compability, you can still create a spatial table in two stages using the management functions.
Create a normal non-spatial table.
For example: CREATE TABLE ROADS ( ID int4, ROAD_NAME varchar(25) )
Add a spatial column to the table using the OpenGIS "AddGeometryColumn" function. Refer to AddGeometryColumn for more details.
The syntax is:
AddGeometryColumn( <schema_name>, <table_name>, <column_name>, <srid>, <type>, <dimension> )
Or, using current schema:
AddGeometryColumn( <table_name>, <column_name>, <srid>, <type>, <dimension> )
Example1: SELECT AddGeometryColumn('public', 'roads', 'geom', 423, 'LINESTRING', 2)
Example2: SELECT AddGeometryColumn( 'roads', 'geom', 423, 'LINESTRING', 2)
Here is an example of SQL used to create a table and add a spatial column (assuming that an SRID of 128 exists already):
CREATE TABLE parks ( park_id INTEGER, park_name VARCHAR, park_date DATE, park_type VARCHAR ); SELECT AddGeometryColumn('parks', 'park_geom', 128, 'MULTIPOLYGON', 2 );
Here is another example, using the generic "geometry" type and the undefined SRID value of 0:
CREATE TABLE roads ( road_id INTEGER, road_name VARCHAR ); SELECT AddGeometryColumn( 'roads', 'roads_geom', 0, 'GEOMETRY', 3 );
The AddGeometryColumn() approach creates a geometry column and also registers the new column in the geometry_columns table. If your software utilizes geometry_columns, then any geometry columns you need to query by must be registered in this view. Starting with PostGIS 2.0, geometry_columns is no longer editable and all geometry columns are autoregistered.
However they may be registered as a generic geometry column if the column was not defined as a specific type during creation.
Two of the cases where this may happen, but you can't use AddGeometryColumn, is in the case of SQL Views and bulk inserts. For these cases, you can correct the registration in the geometry_columns table by constraining the column. Note in PostGIS 2.0+ if your column is typmod based, the creation process would register it correctly, so no need to do anything.
--Lets say you have a view created like this CREATE VIEW public.vwmytablemercator AS SELECT gid, ST_Transform(geom,3395) As geom, f_name FROM public.mytable; -- For it to register correctly in PostGIS 2.0+ -- You need to cast the geometry -- DROP VIEW public.vwmytablemercator; CREATE VIEW public.vwmytablemercator AS SELECT gid, ST_Transform(geom,3395)::geometry(Geometry, 3395) As geom, f_name FROM public.mytable; -- If you know the geometry type for sure is a 2D POLYGON then you could do DROP VIEW public.vwmytablemercator; CREATE VIEW public.vwmytablemercator AS SELECT gid, ST_Transform(geom,3395)::geometry(Polygon, 3395) As geom, f_name FROM public.mytable;
--Lets say you created a derivative table by doing a bulk insert SELECT poi.gid, poi.geom, citybounds.city_name INTO myschema.my_special_pois FROM poi INNER JOIN citybounds ON ST_Intersects(citybounds.geom, poi.geom); --Create 2d index on new table CREATE INDEX idx_myschema_myspecialpois_geom_gist ON myschema.my_special_pois USING gist(geom); -- If your points are 3D points or 3M points, -- then you might want to create an nd index instead of a 2d index -- like so CREATE INDEX my_special_pois_geom_gist_nd ON my_special_pois USING gist(geom gist_geometry_ops_nd); --To manually register this new table's geometry column in geometry_columns -- Note that this approach will work for both PostGIS 2.0+ and PostGIS 1.4+ -- For PostGIS 2.0 it will also change the underlying structure of the table to -- to make the column typmod based. -- For PostGIS prior to 2.0, this technique can also be used to register views SELECT populate_geometry_columns('myschema.my_special_pois'::regclass); --If you are using PostGIS 2.0 and for whatever reason, you -- you need the old constraint based definition behavior -- (such as case of inherited tables where all children do not have the same type and srid) -- set new optional use_typmod argument to false SELECT populate_geometry_columns('myschema.my_special_pois'::regclass, false);
Although the old-constraint based method is still supported, a constraint-based geomentry column used directly in a view, will not register correctly in geometry_columns, as will a typmod one. In this example we define a column using typmod and another using constraints.
CREATE TABLE pois_ny(gid SERIAL PRIMARY KEY , poi_name text, cat varchar(20) , geom geometry(POINT,4326) ); SELECT AddGeometryColumn('pois_ny', 'geom_2160', 2160, 'POINT', 2, false);
If we run in psql
\d pois_ny;
We observe they are defined differently -- one is typmod, one is constraint
Table "public.pois_ny" Column | Type | Modifiers -----------+-----------------------+------------------------------------------------------ gid | integer | not null default nextval('pois_ny_gid_seq'::regclass) poi_name | text | cat | character varying(20) | geom | geometry(Point,4326) | geom_2160 | geometry | Indexes: "pois_ny_pkey" PRIMARY KEY, btree (gid) Check constraints: "enforce_dims_geom_2160" CHECK (st_ndims(geom_2160) = 2) "enforce_geotype_geom_2160" CHECK (geometrytype(geom_2160) = 'POINT'::text OR geom_2160 IS NULL) "enforce_srid_geom_2160" CHECK (st_srid(geom_2160) = 2160)
In geometry_columns, they both register correctly
SELECT f_table_name, f_geometry_column, srid, type FROM geometry_columns WHERE f_table_name = 'pois_ny';
f_table_name | f_geometry_column | srid | type -------------+-------------------+------+------- pois_ny | geom | 4326 | POINT pois_ny | geom_2160 | 2160 | POINT
However -- if we were to create a view like this
CREATE VIEW vw_pois_ny_parks AS SELECT * FROM pois_ny WHERE cat='park'; SELECT f_table_name, f_geometry_column, srid, type FROM geometry_columns WHERE f_table_name = 'vw_pois_ny_parks';
The typmod based geom view column registers correctly, but the constraint based one does not.
f_table_name | f_geometry_column | srid | type ------------------+-------------------+------+---------- vw_pois_ny_parks | geom | 4326 | POINT vw_pois_ny_parks | geom_2160 | 0 | GEOMETRY
This may change in future versions of PostGIS, but for now To force the constraint based view column to register correctly, we need to do this:
DROP VIEW vw_pois_ny_parks; CREATE VIEW vw_pois_ny_parks AS SELECT gid, poi_name, cat , geom , geom_2160::geometry(POINT,2160) As geom_2160 FROM pois_ny WHERE cat='park'; SELECT f_table_name, f_geometry_column, srid, type FROM geometry_columns WHERE f_table_name = 'vw_pois_ny_parks';
f_table_name | f_geometry_column | srid | type ------------------+-------------------+------+------- vw_pois_ny_parks | geom | 4326 | POINT vw_pois_ny_parks | geom_2160 | 2160 | POINT
PostGIS is compliant with the Open Geospatial Consortium’s (OGC) OpenGIS Specifications. As such, many PostGIS methods require, or more accurately, assume that geometries that are operated on are both simple and valid. For example, it does not make sense to calculate the area of a polygon that has a hole defined outside of the polygon, or to construct a polygon from a non-simple boundary line.
According to the OGC Specifications, a simple geometry is one that has no anomalous geometric points, such as self intersection or self tangency and primarily refers to 0 or 1-dimensional geometries (i.e. [MULTI]POINT, [MULTI]LINESTRING
). Geometry validity, on the other hand, primarily refers to 2-dimensional geometries (i.e. [MULTI]POLYGON)
and defines the set of assertions that characterizes a valid polygon. The description of each geometric class includes specific conditions that further detail geometric simplicity and validity.
A POINT
is inheritably simple as a 0-dimensional geometry object.
MULTIPOINT
s are simple if no two coordinates (POINT
s) are equal (have identical coordinate values).
A LINESTRING
is simple if it does not pass through the same POINT
twice (except for the endpoints, in which case it is referred to as a linear ring and additionally considered closed).
(a) and (c) are simple |
A MULTILINESTRING
is simple only if all of its elements are simple and the only intersection between any two elements occurs at POINT
s that are on the boundaries of both elements.
(e) and (f) are simple |
By definition, a POLYGON
is always simple. It is valid if no two rings in the boundary (made up of an exterior ring and interior rings) cross. The boundary of a POLYGON
may intersect at a POINT
but only as a tangent (i.e. not on a line). A POLYGON
may not have cut lines or spikes and the interior rings must be contained entirely within the exterior ring.
(h) and (i) are valid |
A MULTIPOLYGON
is valid if and only if all of its elements are valid and the interiors of no two elements intersect. The boundaries of any two elements may touch, but only at a finite number of POINT
s.
(n) and (o) are not valid |
Most of the functions implemented by the GEOS library rely on the assumption that your geometries are valid as specified by the OpenGIS Simple Feature Specification. To check simplicity or validity of geometries you can use the ST_IsSimple() and ST_IsValid()
-- Typically, it doesn't make sense to check -- for validity on linear features since it will always return TRUE. -- But in this example, PostGIS extends the definition of the OGC IsValid -- by returning false if a LineString has less than 2 *distinct* vertices. gisdb=# SELECT ST_IsValid('LINESTRING(0 0, 1 1)'), ST_IsValid('LINESTRING(0 0, 0 0, 0 0)'); st_isvalid | st_isvalid ------------+----------- t | f
By default, PostGIS does not apply this validity check on geometry input, because testing for validity needs lots of CPU time for complex geometries, especially polygons. If you do not trust your data sources, you can manually enforce such a check to your tables by adding a check constraint:
ALTER TABLE mytable ADD CONSTRAINT geometry_valid_check CHECK (ST_IsValid(the_geom));
If you encounter any strange error messages such as "GEOS Intersection() threw an error!" or "JTS Intersection() threw an error!" when calling PostGIS functions with valid input geometries, you likely found an error in either PostGIS or one of the libraries it uses, and you should contact the PostGIS developers. The same is true if a PostGIS function returns an invalid geometry for valid input.
Strictly compliant OGC geometries cannot have Z or M values. The ST_IsValid() function won't consider higher dimensioned geometries invalid! Invocations of AddGeometryColumn() will add a constraint checking geometry dimensions, so it is enough to specify 2 there. |
It is sometimes the case that the typical spatial predicates (ST_Contains, ST_Crosses, ST_Intersects, ST_Touches, ...) are insufficient in and of themselves to adequately provide that desired spatial filter.
For example, consider a linear dataset representing a road network. It may be the task of a GIS analyst to identify all road segments that cross each other, not at a point, but on a line, perhaps invalidating some business rule. In this case, ST_Crosses does not adequately provide the necessary spatial filter since, for linear features, it returns One two-step solution might be to first perform the actual intersection (ST_Intersection) of pairs of road segments that spatially intersect (ST_Intersects), and then compare the intersection's ST_GeometryType with ' A more elegant / faster solution may indeed be desirable. |
A second [theoretical] example may be that of a GIS analyst trying to locate all wharfs or docks that intersect a lake's boundary on a line and where only one end of the wharf is up on shore. In other words, where a wharf is within, but not completely within a lake, intersecting the boundary of a lake on a line, and where the wharf's endpoints are both completely within and on the boundary of the lake. The analyst may need to use a combination of spatial predicates to isolate the sought after features:
|
So enters the Dimensionally Extended 9 Intersection Model, or DE-9IM for short.
According to the OpenGIS Simple Features Implementation Specification for SQL, "the basic approach to comparing two geometries is to make pair-wise tests of the intersections between the Interiors, Boundaries and Exteriors of the two geometries and to classify the relationship between the two geometries based on the entries in the resulting 'intersection' matrix."
The boundary of a geometry is the set of geometries of the next lower dimension. For POINT
s, which have a dimension of 0, the boundary is the empty set. The boundary of a LINESTRING
are the two endpoints. For POLYGON
s, the boundary is the linework that make up the exterior and interior rings.
The interior of a geometry are those points of a geometry that are left when the boundary is removed. For POINT
s, the interior is the POINT
itself. The interior of a LINESTRING
are the set of real points between the endpoints. For POLYGON
s, the interior is the areal surface inside the polygon.
The exterior of a geometry is the universe, an areal surface, not on the interior or boundary of the geometry.
Given geometry a, where the I(a), B(a), and E(a) are the Interior, Boundary, and Exterior of a, the mathematical representation of the matrix is:
Interior | Boundary | Exterior | |
---|---|---|---|
Interior | dim( I(a) ∩ I(b) ) | dim( I(a) ∩ B(b) ) | dim( I(a) ∩ E(b) ) |
Boundary | dim( B(a) ∩ I(b) ) | dim( B(a) ∩ B(b) ) | dim( B(a) ∩ E(b) ) |
Exterior | dim( E(a) ∩ I(b) ) | dim( E(a) ∩ B(b) ) | dim( E(a) ∩ E(b) ) |
Where dim(a) is the dimension of a as specified by ST_Dimension but has the domain of {0,1,2,T,F,*}
0
=> point
1
=> line
2
=> area
T
=> {0,1,2}
F
=> empty set
*
=> don't care
Visually, for two overlapping polygonal geometries, this looks like:
|
Read from left to right and from top to bottom, the dimensional matrix is represented, '212101212'.
A relate matrix that would therefore represent our first example of two lines that intersect on a line would be: '1*1***1**'
-- Identify road segments that cross on a line SELECT a.id FROM roads a, roads b WHERE a.id != b.id AND a.geom && b.geom AND ST_Relate(a.geom, b.geom, '1*1***1**');
A relate matrix that represents the second example of wharfs partly on the lake's shoreline would be '102101FF2'
-- Identify wharfs partly on a lake's shoreline SELECT a.lake_id, b.wharf_id FROM lakes a, wharfs b WHERE a.geom && b.geom AND ST_Relate(a.geom, b.geom, '102101FF2');
For more information or reading, see:
OpenGIS Simple Features Implementation Specification for SQL (version 1.1, section 2.1.13.2)
Encyclopedia of GIS By Hui Xiong
Once you have created a spatial table, you are ready to upload GIS data to the database. Currently, there are two ways to get data into a PostGIS/PostgreSQL database: using formatted SQL statements or using the Shape file loader/dumper.
If you can convert your data to a text representation, then using formatted SQL might be the easiest way to get your data into PostGIS. As with Oracle and other SQL databases, data can be bulk loaded by piping a large text file full of SQL "INSERT" statements into the SQL terminal monitor.
A data upload file (roads.sql
for example) might look like this:
BEGIN; INSERT INTO roads (road_id, roads_geom, road_name) VALUES (1,ST_GeomFromText('LINESTRING(191232 243118,191108 243242)',-1),'Jeff Rd'); INSERT INTO roads (road_id, roads_geom, road_name) VALUES (2,ST_GeomFromText('LINESTRING(189141 244158,189265 244817)',-1),'Geordie Rd'); INSERT INTO roads (road_id, roads_geom, road_name) VALUES (3,ST_GeomFromText('LINESTRING(192783 228138,192612 229814)',-1),'Paul St'); INSERT INTO roads (road_id, roads_geom, road_name) VALUES (4,ST_GeomFromText('LINESTRING(189412 252431,189631 259122)',-1),'Graeme Ave'); INSERT INTO roads (road_id, roads_geom, road_name) VALUES (5,ST_GeomFromText('LINESTRING(190131 224148,190871 228134)',-1),'Phil Tce'); INSERT INTO roads (road_id, roads_geom, road_name) VALUES (6,ST_GeomFromText('LINESTRING(198231 263418,198213 268322)',-1),'Dave Cres'); COMMIT;
The data file can be piped into PostgreSQL very easily using the "psql" SQL terminal monitor:
psql -d [database] -f roads.sql
The shp2pgsql
data loader converts ESRI Shape files into SQL suitable for insertion into a PostGIS/PostgreSQL database either in geometry or geography format. The loader has several operating modes distinguished by command line flags:
In addition to the shp2pgsql command-line loader, there is an shp2pgsql-gui
graphical interface with most of the options as the command-line loader, but may be easier to use for one-off non-scripted loading or if you are new to PostGIS. It can also be configured as a plugin to PgAdminIII.
Creates a new table and populates it from the shapefile. This is the default mode.
Appends data from the Shape file into the database table. Note that to use this option to load multiple files, the files must have the same attributes and same data types.
Drops the database table before creating a new table with the data in the Shape file.
Only produces the table creation SQL code, without adding any actual data. This can be used if you need to completely separate the table creation and data loading steps.
Display help screen.
Use the PostgreSQL "dump" format for the output data. This can be combined with -a, -c and -d. It is much faster to load than the default "insert" SQL format. Use this for very large data sets.
Creates and populates the geometry tables with the specified SRID. Optionally specifies that the input shapefile uses the given FROM_SRID, in which case the geometries will be reprojected to the target SRID. FROM_SRID cannot be specified with -D.
Keep identifiers' case (column, schema and attributes). Note that attributes in Shapefile are all UPPERCASE.
Coerce all integers to standard 32-bit integers, do not create 64-bit bigints, even if the DBF header signature appears to warrant it.
Create a GiST index on the geometry column.
-m a_file_name
Specify a file containing a set of mappings of (long) column names to 10 character DBF column names. The content of the file is one or more lines of two names separated by white space and no trailing or leading space. For example:
COLUMNNAME DBFFIELD1 AVERYLONGCOLUMNNAME DBFFIELD2
Generate simple geometries instead of MULTI geometries. Will only succeed if all the geometries are actually single (I.E. a MULTIPOLYGON with a single shell, or or a MULTIPOINT with a single vertex).
Force the output geometry to have the specified dimensionality. Use the following strings to indicate the dimensionality: 2D, 3DZ, 3DM, 4D.
If the input has fewer dimensions that specified, the output will have those dimensions filled in with zeroes. If the input has more dimensions that specified, the unwanted dimensions will be stripped.
Output WKT format, instead of WKB. Note that this can introduce coordinate drifts due to loss of precision.
Execute each statement on its own, without using a transaction. This allows loading of the majority of good data when there are some bad geometries that generate errors. Note that this cannot be used with the -D flag as the "dump" format always uses a transaction.
Specify encoding of the input data (dbf file). When used, all attributes of the dbf are converted from the specified encoding to UTF8. The resulting SQL output will contain a SET CLIENT_ENCODING to UTF8
command, so that the backend will be able to reconvert from UTF8 to whatever encoding the database is configured to use internally.
NULL geometries handling policy (insert*,skip,abort)
-n Only import DBF file. If your data has no corresponding shapefile, it will automatically switch to this mode and load just the dbf. So setting this flag is only needed if you have a full shapefile set, and you only want the attribute data and no geometry.
Use geography type instead of geometry (requires lon/lat data) in WGS84 long lat (SRID=4326)
Specify the tablespace for the new table. Indexes will still use the default tablespace unless the -X parameter is also used. The PostgreSQL documentation has a good description on when to use custom tablespaces.
Specify the tablespace for the new table's indexes. This applies to the primary key index, and the GIST spatial index if -I is also used.
An example session using the loader to create an input file and uploading it might look like this:
# shp2pgsql -c -D -s 4269 -i -I shaperoads.shp myschema.roadstable > roads.sql # psql -d roadsdb -f roads.sql
A conversion and upload can be done all in one step using UNIX pipes:
# shp2pgsql shaperoads.shp myschema.roadstable | psql -d roadsdb
Data can be extracted from the database using either SQL or the Shape file loader/dumper. In the section on SQL we will discuss some of the operators available to do comparisons and queries on spatial tables.
The most straightforward means of pulling data out of the database is to use a SQL select query to reduce the number of RECORDS and COLUMNS returned and dump the resulting columns into a parsable text file:
db=# SELECT road_id, ST_AsText(road_geom) AS geom, road_name FROM roads; road_id | geom | road_name --------+-----------------------------------------+----------- 1 | LINESTRING(191232 243118,191108 243242) | Jeff Rd 2 | LINESTRING(189141 244158,189265 244817) | Geordie Rd 3 | LINESTRING(192783 228138,192612 229814) | Paul St 4 | LINESTRING(189412 252431,189631 259122) | Graeme Ave 5 | LINESTRING(190131 224148,190871 228134) | Phil Tce 6 | LINESTRING(198231 263418,198213 268322) | Dave Cres 7 | LINESTRING(218421 284121,224123 241231) | Chris Way (6 rows)
However, there will be times when some kind of restriction is necessary to cut down the number of fields returned. In the case of attribute-based restrictions, just use the same SQL syntax as normal with a non-spatial table. In the case of spatial restrictions, the following operators are available/useful:
This operator tells whether the bounding box of one geometry intersects the bounding box of another.
This tests whether two geometries are geometrically identical. For example, if 'POLYGON((0 0,1 1,1 0,0 0))' is the same as 'POLYGON((0 0,1 1,1 0,0 0))' (it is).
This operator is a little more naive, it only tests whether the bounding boxes of two geometries are the same.
Next, you can use these operators in queries. Note that when specifying geometries and boxes on the SQL command line, you must explicitly turn the string representations into geometries by using the "ST_GeomFromText()" function. The 312 is a fictitious spatial reference system that matches our data. So, for example:
SELECT road_id, road_name FROM roads WHERE ST_OrderingEquals(roads_geom , ST_GeomFromText('LINESTRING(191232 243118,191108 243242)',312) ) ;
The above query would return the single record from the "ROADS_GEOM" table in which the geometry was equal to that value.
When using the "&&" operator, you can specify either a BOX3D as the comparison feature or a GEOMETRY. When you specify a GEOMETRY, however, its bounding box will be used for the comparison.
SELECT road_id, road_name FROM roads WHERE roads_geom && ST_GeomFromText('POLYGON((...))',312);
The above query will use the bounding box of the polygon for comparison purposes.
The most common spatial query will probably be a "frame-based" query, used by client software, like data browsers and web mappers, to grab a "map frame" worth of data for display. Using a "BOX3D" object for the frame, such a query looks like this:
SELECT ST_AsText(roads_geom) AS geom FROM roads WHERE roads_geom && ST_MakeEnvelope(191232, 243117,191232, 243119,312);
Note the use of the SRID 312, to specify the projection of the envelope.
The pgsql2shp
table dumper connects directly to the database and converts a table (possibly defined by a query) into a shape file. The basic syntax is:
pgsql2shp [<options>] <database> [<schema>.]<table>
pgsql2shp [<options>] <database> <query>
The commandline options are:
Write the output to a particular filename.
The database host to connect to.
The port to connect to on the database host.
The password to use when connecting to the database.
The username to use when connecting to the database.
In the case of tables with multiple geometry columns, the geometry column to use when writing the shape file.
Use a binary cursor. This will make the operation faster, but will not work if any NON-geometry attribute in the table lacks a cast to text.
Raw mode. Do not drop the gid
field, or escape column names.
For backward compatibility: write a 3-dimensional shape file when dumping from old (pre-1.0.0) postgis databases (the default is to write a 2-dimensional shape file in that case). Starting from postgis-1.0.0+, dimensions are fully encoded.
filename
Remap identifiers to ten character names. The content of the file is lines of two symbols separated by a single white space and no trailing or leading space: VERYLONGSYMBOL SHORTONE ANOTHERVERYLONGSYMBOL SHORTER etc.
Indexes are what make using a spatial database for large data sets possible. Without indexing, any search for a feature would require a "sequential scan" of every record in the database. Indexing speeds up searching by organizing the data into a search tree which can be quickly traversed to find a particular record. PostgreSQL supports three kinds of indexes by default: B-Tree indexes, R-Tree indexes, and GiST indexes.
B-Trees are used for data which can be sorted along one axis; for example, numbers, letters, dates. GIS data cannot be rationally sorted along one axis (which is greater, (0,0) or (0,1) or (1,0)?) so B-Tree indexing is of no use for us.
R-Trees break up data into rectangles, and sub-rectangles, and sub-sub rectangles, etc. R-Trees are used by some spatial databases to index GIS data, but the PostgreSQL R-Tree implementation is not as robust as the GiST implementation.
GiST (Generalized Search Trees) indexes break up data into "things to one side", "things which overlap", "things which are inside" and can be used on a wide range of data-types, including GIS data. PostGIS uses an R-Tree index implemented on top of GiST to index GIS data.
GiST stands for "Generalized Search Tree" and is a generic form of indexing. In addition to GIS indexing, GiST is used to speed up searches on all kinds of irregular data structures (integer arrays, spectral data, etc) which are not amenable to normal B-Tree indexing.
Once a GIS data table exceeds a few thousand rows, you will want to build an index to speed up spatial searches of the data (unless all your searches are based on attributes, in which case you'll want to build a normal index on the attribute fields).
The syntax for building a GiST index on a "geometry" column is as follows:
CREATE INDEX [indexname] ON [tablename] USING GIST ( [geometryfield] );
The above syntax will always build a 2D-index. To get the an n-dimensional index supported in PostGIS 2.0+ for the geometry type, you can create one using this syntax
CREATE INDEX [indexname] ON [tablename] USING GIST ([geometryfield] gist_geometry_ops_nd);
Building a spatial index is a computationally intensive exercise: on tables of around 1 million rows, on a 300MHz Solaris machine, we have found building a GiST index takes about 1 hour. After building an index, it is important to force PostgreSQL to collect table statistics, which are used to optimize query plans:
VACUUM ANALYZE [table_name] [(column_name)]; -- This is only needed for PostgreSQL 7.4 installations and below SELECT UPDATE_GEOMETRY_STATS([table_name], [column_name]);
GiST indexes have two advantages over R-Tree indexes in PostgreSQL. Firstly, GiST indexes are "null safe", meaning they can index columns which include null values. Secondly, GiST indexes support the concept of "lossiness" which is important when dealing with GIS objects larger than the PostgreSQL 8K page size. Lossiness allows PostgreSQL to store only the "important" part of an object in an index -- in the case of GIS objects, just the bounding box. GIS objects larger than 8K will cause R-Tree indexes to fail in the process of being built.
Ordinarily, indexes invisibly speed up data access: once the index is built, the query planner transparently decides when to use index information to speed up a query plan. Unfortunately, the PostgreSQL query planner does not optimize the use of GiST indexes well, so sometimes searches which should use a spatial index instead default to a sequence scan of the whole table.
If you find your spatial indexes are not being used (or your attribute indexes, for that matter) there are a couple things you can do:
Firstly, make sure statistics are gathered about the number and distributions of values in a table, to provide the query planner with better information to make decisions around index usage. For PostgreSQL 7.4 installations and below this is done by running update_geometry_stats([table_name, column_name]) (compute distribution) and VACUUM ANALYZE [table_name] [column_name] (compute number of values). Starting with PostgreSQL 8.0 running VACUUM ANALYZE will do both operations. You should regularly vacuum your databases anyways -- many PostgreSQL DBAs have VACUUM run as an off-peak cron job on a regular basis.
If vacuuming does not work, you can force the planner to use the index information by using the SET ENABLE_SEQSCAN=OFF command. You should only use this command sparingly, and only on spatially indexed queries: generally speaking, the planner knows better than you do about when to use normal B-Tree indexes. Once you have run your query, you should consider setting ENABLE_SEQSCAN
back on, so that other queries will utilize the planner as normal.
As of version 0.6, it should not be necessary to force the planner to use the index with |
If you find the planner wrong about the cost of sequential vs index scans try reducing the value of random_page_cost in postgresql.conf or using SET random_page_cost=#. Default value for the parameter is 4, try setting it to 1 or 2. Decrementing the value makes the planner more inclined of using Index scans.
The raison d'etre of spatial database functionality is performing queries inside the database which would ordinarily require desktop GIS functionality. Using PostGIS effectively requires knowing what spatial functions are available, and ensuring that appropriate indexes are in place to provide good performance. The SRID of 312 used in these examples is purely for demonstration. You should be using a REAL SRID listed in the the spatial_ref_sys table and one that matches the projection of your data. If your data has no spatial reference system specified, you should be THINKING very thoughtfully why it doesn't and maybe it should. If your reason is because you are modeling something that doesn't have a geographic spatial reference system defined such as the internals of a molecule or a good location on Mars to transport the human race in the event of a nuclear holocaust, then simply leave out the SRID or make one up and insert it in the spatial_ref_sys
table.
When constructing a query it is important to remember that only the bounding-box-based operators such as && can take advantage of the GiST spatial index. Functions such as ST_Distance()
cannot use the index to optimize their operation. For example, the following query would be quite slow on a large table:
SELECT the_geom FROM geom_table WHERE ST_Distance(the_geom, ST_GeomFromText('POINT(100000 200000)', 312)) < 100
This query is selecting all the geometries in geom_table which are within 100 units of the point (100000, 200000). It will be slow because it is calculating the distance between each point in the table and our specified point, ie. one ST_Distance()
calculation for each row in the table. We can avoid this by using the && operator to reduce the number of distance calculations required:
SELECT the_geom FROM geom_table WHERE ST_DWithin(the_geom, ST_MakeEnvelope(90900, 190900, 100100, 200100,312), 100)
This query selects the same geometries, but it does it in a more efficient way. Assuming there is a GiST index on the_geom, the query planner will recognize that it can use the index to reduce the number of rows before calculating the result of the ST_distance()
function. Notice that the ST_MakeEnvelope
geometry which is used in the && operation is a 200 unit square box centered on the original point - this is our "query box". The && operator uses the index to quickly reduce the result set down to only those geometries which have bounding boxes that overlap the "query box". Assuming that our query box is much smaller than the extents of the entire geometry table, this will drastically reduce the number of distance calculations that need to be done.
Change in Behavior | |
---|---|
As of PostGIS 1.3.0, most of the Geometry Relationship Functions, with the notable exceptions of ST_Disjoint and ST_Relate, include implicit bounding box overlap operators. |
The examples in this section will make use of two tables, a table of linear roads, and a table of polygonal municipality boundaries. The table definitions for the bc_roads
table is:
Column | Type | Description ------------+-------------------+------------------- gid | integer | Unique ID name | character varying | Road Name the_geom | geometry | Location Geometry (Linestring)
The table definition for the bc_municipality
table is:
Column | Type | Description -----------+-------------------+------------------- gid | integer | Unique ID code | integer | Unique ID name | character varying | City / Town Name the_geom | geometry | Location Geometry (Polygon)
Table of Contents
Nella maggior parte dei casi, creerete i raster PostGIS caricando file esterni tramite il raster loader raster2pgsql
compreso nell'installazione.
raster2pgsql
è un loader raster che trasforma i formati raster supportati da GDAL in formato SQL utile al caricamento in una tabella raster di PostGIS. Può caricare anche cartelle di file, e creare delle anteprime.
Dato che in genere raster2pgsql viene compilato come parte di PostGIS (a meno che non abbiate compilato una vostra libreria GDAL a parte), i tipi di raster supportati dall'eseguibile saranno gli stessi che sono compilati per GDAL. Per avere una lista dei tipi di raster supportati dal vostro raster2pgsql utilizzate l'opzione -G
. Questi tipi dovrebbero essere gli stessi disponibilii tramite la vostra installazione di PostGIS e documentati qui ST_GDALDrivers, se usate la stessa libreria gdal per entrambi i programmi.
La vecchia versione di questo strumento era uno script in python. L'eseguibile sostituisce lo script in python. Se vi trovate ad avere bisogno dello script in Python, alcuni esempio sono reperibili alla pagina GDAL PostGIS Raster Driver Usage. Notate che lo script Python raster2pgsql potrebbe non funzionare con le future versioni di raster PostGIS e che non è più supportato. |
When creating overviews of a specific factor from a set of rasters that are aligned, it is possible for the overviews to not align. Visit http://trac.osgeo.org/postgis/ticket/1764 for an example where the overviews do not align. |
ESEMPIO DI UTILIZZO:
raster2pgsqlopzioni raster
nome file raster
snome schema
.nome tabella
> out.sql
Mostra una schermata di aiuto. L'aiuto verra mostrato inoltre se non assegnate alcun parametro.
Elenca i formati raster supportati.
Crea una nuova tabella e carica in questa il/i raster. Questa è la modalita di default
Accoda il o i raster a una tabella esistente.
Elimina la tabella, ne crea una nuova e vi carica il/i dati raster
Modalità di preparazione. Crea solamente la tabella.
Apply raster constraints -- srid, pixelsize etc. to ensure raster is properly registered in raster_columns
view.
Disable setting the max extent constraint. Only applied if -C flag is also used.
Set the constraints (spatially unique and coverage tile) for regular blocking. Only applied if -C flag is also used.
Assegna lo SRID specificato al raster in uscita. Se non fornito o uguale a zero, saranno controllati i metadati del raster per determinare uno SRID appropriato.
Indice (a partire da 1) della banda da estrarre dal raster. Per specificare più di un indice di banda, separare con una virgola (,). Se non specificato, saranno estratte tutte le bande.
Cut raster into tiles to be inserted one per table row. TILE_SIZE
is expressed as WIDTHxHEIGHT or set to the value "auto" to allow the loader to compute an appropriate tile size using the first raster and applied to all rasters.
Pad right-most and bottom-most tiles to guarantee that all tiles have the same width and height.
Register the raster as a filesystem (out-db) raster.
Nel database vengono salvati solo i metadati e il percorso del raster (non i pixel).
OVERVIEW_FACTOR
Create overview of the raster. For more than one factor, separate with comma(,). Overview table name follows the pattern o_overview factor
_table
, where overview factor
is a placeholder for numerical overview factor and table
is replaced with the base table name. Created overview is stored in the database and is not affected by -R. Note that your generated sql file will contain both the main table and overview tables.
NODATA
Valore da usare come NODATA per le bande senza un valore NODATA.
Wrap PostgreSQL identifiers in quotes
Specifica il nome della colonna di destinazione per i raster. Il default è 'rast'.
Aggiunge una colonna con il nome del file
Specify the name of the filename column. Implies -F.
Wrap PostgreSQL identifiers in quotes.
Crea un indice GiST sulla colonna raster.
Vacuum analyze the raster table.
Skip NODATA value checks for each raster band.
tablespace
Specificare il tablespace per la nuova tabella. Notare che gli indici (compresa la chiave primaria) useranno sempre il tablespace di default, a meno che non venga usato anche il flag -X.
tablespace
Specifica il tablespace per il nuovo indice della tabella. Si applica alla chiave primaria e all'indice spaziale se viene usato il flag -I
Utilizza comandi di copia anziché di inserimento.
Esegui ogni comando individualmente, non utilizzare una transazione.
Controlla l'ordine dei byte prodotti nell'output binario del raster: specificare 0 per XDR e 1 per NDR (il default). Al momento viene supportato solo lo NDR.
versione
Specifica la versione del formato in uscita. Il default è 0. Al momento 0 è l'unico supportato.
Una sessione di esempio che utilizzi il loader per creare un file di input e per caricarlo a pezzi di tile 100x100 potrebbe essere il seguente:
Potete omettere il nome dello schema, per esempio |
raster2pgsql -s 4326 -I -C -M *.tif -F -t 100x100 public.demelevation > elev.sql psql -d gisdb -f elev.sql
La conversione e il caricamento possono essere eseguiti in un unico passaggio tramite le pipe UNIX:
raster2pgsql -s 4326 -I -C -M *.tif -F -t 100x100 public.demelevation | psql -d gisdb
Load rasters Massachusetts state plane meters aerial tiles into a schema called aerial
and create a full view, 2 and 4 level overview tables, use copy mode for inserting (no intermediary file just straight to db), and -e don't force everything in a transaction (good if you want to see data in tables right away without waiting). Break up the rasters into 128x128 pixel tiles and apply raster constraints. Use copy mode instead of table insert. (-F) Include a field called filename to hold the name of the file the tiles were cut from.
raster2pgsql -I -C -e -Y -F -s 26986 -t 128x128 -l 2,4 bostonaerials2008/*.jpg aerials.boston | psql -U postgres -d gisdb -h localhost -p 5432
--ottieni la lista dei tipi raster supportati: raster2pgsql -G
Il parametro -G restituirà un elenco tipo
Available GDAL raster formats: Virtual Raster GeoTIFF National Imagery Transmission Format Raster Product Format TOC format ECRG TOC format Erdas Imagine Images (.img) CEOS SAR Image CEOS Image JAXA PALSAR Product Reader (Level 1.1/1.5) Ground-based SAR Applications Testbed File Format (.gff) ELAS Arc/Info Binary Grid Arc/Info ASCII Grid GRASS ASCII Grid SDTS Raster DTED Elevation Raster Portable Network Graphics JPEG JFIF In Memory Raster Japanese DEM (.mem) Graphics Interchange Format (.gif) Graphics Interchange Format (.gif) Envisat Image Format Maptech BSB Nautical Charts X11 PixMap Format MS Windows Device Independent Bitmap SPOT DIMAP AirSAR Polarimetric Image RadarSat 2 XML Product PCIDSK Database File PCRaster Raster File ILWIS Raster Map SGI Image File Format 1.0 SRTMHGT File Format Leveller heightfield Terragen heightfield USGS Astrogeology ISIS cube (Version 3) USGS Astrogeology ISIS cube (Version 2) NASA Planetary Data System EarthWatch .TIL ERMapper .ers Labelled NOAA Polar Orbiter Level 1b Data Set FIT Image GRIdded Binary (.grb) Raster Matrix Format EUMETSAT Archive native (.nat) Idrisi Raster A.1 Intergraph Raster Golden Software ASCII Grid (.grd) Golden Software Binary Grid (.grd) Golden Software 7 Binary Grid (.grd) COSAR Annotated Binary Matrix (TerraSAR-X) TerraSAR-X Product DRDC COASP SAR Processor Raster R Object Data Store Portable Pixmap Format (netpbm) USGS DOQ (Old Style) USGS DOQ (New Style) ENVI .hdr Labelled ESRI .hdr Labelled Generic Binary (.hdr Labelled) PCI .aux Labelled Vexcel MFF Raster Vexcel MFF2 (HKV) Raster Fuji BAS Scanner Image GSC Geogrid EOSAT FAST Format VTP .bt (Binary Terrain) 1.3 Format Erdas .LAN/.GIS Convair PolGASP Image Data and Analysis NLAPS Data Format Erdas Imagine Raw DIPEx FARSITE v.4 Landscape File (.lcp) NOAA Vertical Datum .GTX NADCON .los/.las Datum Grid Shift NTv2 Datum Grid Shift ACE2 Snow Data Assimilation System Swedish Grid RIK (.rik) USGS Optional ASCII DEM (and CDED) GeoSoft Grid Exchange Format Northwood Numeric Grid Format .grd/.tab Northwood Classified Grid Format .grc/.tab ARC Digitized Raster Graphics Standard Raster Product (ASRP/USRP) Magellan topo (.blx) SAGA GIS Binary Grid (.sdat) Kml Super Overlay ASCII Gridded XYZ HF2/HFZ heightfield raster OziExplorer Image File USGS LULC Composite Theme Grid Arc/Info Export E00 GRID ZMap Plus Grid NOAA NGS Geoid Height Grids
In varie occasioni vorrete creare raster e tabelle raster direttamente nel database. Per questo esiste una pletora di funzioni. Questi sono i passi generali da seguire.
Creare una tabella con una colonna raster per contenere i nuovi valori raster può essere ottenuto da:
CREATE TABLE myrasters(rid serial primary key, rast raster);
Esistono molte funzioni per assitervi verso questo obiettivo. Se state creando un raster non derivato da altri raster, inizierete con: ST_MakeEmptyRaster, seguito da ST_AddBand
Potete anche creare raster a partire dalle geometria. Per questo userete ST_AsRaster, magari accompagnato da altre funzioni come ST_Union o ST_MapAlgebraFct, o qualsiasi altra delle funzioni di algebra sulle mappe.
Vi sono poi ancora più opzioni per creare nuove tabelle raster a partire da tabelle esistenti. Per esempio potete creare una tabella raster in una proiezione diversa da una tabella esistente usando ST_Transform
Una volta inseriti dei valori iniziali nella tabella, vorrete creare un indice spaziale sulla colonna raster, con un comando tipo:
CREATE INDEX myrasters_rast_st_convexhull_idx ON myrasters USING gist( ST_ConvexHull(rast) );
Note the use of ST_ConvexHull since most raster operators are based on the convex hull of the rasters.
Pre-2.0 versions of PostGIS raster were based on the envelop rather than the convex hull. For the spatial indexes to work properly you'll need to drop those and replace with convex hull based index. |
Apply raster constraints using AddRasterConstraints
Esistono due tipi di viste dei cataloghi raster fornite con PostGIS. Entrambe le viste utilizzano informazioni contenute nei dati sui limiti delle tabelle raster. Da ciò risulta che le viste dei cataloghi sono sempre consistenti con i dati raster nelle tabelle, dato che sono considerati i limiti di queste.
raster_columns
questa vista riporta il catalogo di tutte le colonne raster nel vostro database.
raster_overviews
questa vista elenca tutte le colonne raste di tabelle nel vostro database che sono utilizzate come vista d'insieme per una tabella con maggiori dettagli. Le tabelle di questo tipo sono generate quando utilizzate l'opzione -l
durante il caricamento.
raster_columns
è un catalogo di tutte le colonne di tabelle che nel vostro database sono di tipo raster. E' una vista che utilizza i limiti applicati alle tabelle, per cui le informazioni sono sempre congruenti, anche se ripristinate una tabella raster dal backup di un altro database. Il catalogo raster_columns
contiene le seguenti colonne.
Se non avete creato le tabelle con il loader o vi siete dimenticati di specificare l'opzione -C
durante il caricamento, potete far applicare i limiti a cose fatte utilizzando AddRasterConstraints, di modo che il catalogo raster_columns
registri le informazioni sulle vostre tile raster.
r_table_catalog
Il database in cui si trova la tabella. Conterrà sempre il database corrente.
r_table_schema
Lo schema database cui la tabella raster appartiene.
r_table_name
tabella raster
r_raster_column
la colonna nella tabella r_table_name
che è di tipo raster. Nulla in PostGIS vi impedisce di avere più colonne raster per tabella, per cui è possibile avere una tabella raster elencata più volte con il riferimento ogni volta a una colonna raster differente..
srid
L'identificatore del sistema di riferimento spaziale del raster. Dovrebbe essere una voce di Section 4.3.1, “The SPATIAL_REF_SYS Table and Spatial Reference Systems”.
scale_x
Il rapporto di scala tra le coordinate geometriche e il pixel, disponibile solo se tutte le tile nella colonna raster hanno lo stesso valore di scale_x
e questo limite è applicato. Si rimanda è ST_ScaleX per ulteriori dettagli.
scale_y
Il rapporto di scala tra le coordinate geometriche e il pixel, disponibile solo se tutte le tile nella colonna raster hanno lo stesso valore di scale_y
e questo limite è applicato. Si rimanda è ST_ScaleY per ulteriori dettagli.
blocksize_x
La larghezza (come numero di pixel in orizzontale) di ogni tile raster. Si rimanda a ST_Width per ulteriori dettagli.
blocksize_y
L'altezza (number of pixels in verticale) di ogni tile raster. Si rimanda a ST_Height per ulteriori dettagli.
stesso_allineamento
Variabile booleana pari a "vero" se tutte le tile raster hanno lo stesso allineamento. Si rimanda a ST_SameAlignment per ulteriori dettagli.
regular_blocking
If the raster column has the spatially unique and coverage tile constraints, the value with be TRUE. Otherwise, it will be FALSE.
numero_bande
Il numero delle bande in ogni tile del set di raster. Questa è la stessa informazione fornita da ST_NumBands
pixel_types
Un vettore che definisce il tipo di pixel per ciascuna banda. In questo vettore avrete un numero di elementi pari al numero delle bande. I valori di pixel_types possono essere tra quelli definiti in ST_BandPixelType.
nodata_values
Un vettore in doppia precisione che spefica i valori nodata_value
per ciascuna banda. Avrete in questo vettore un numero di elementi pari al numero di bande. Questi numeri definiscono il valore del pixel che per ciascuna banda deve essere ignorato nella maggior parte delle operazioni. L'informazion è simile a quella fornita da ST_BandNoDataValue.
out_db
An array of boolean flags indicating if the raster bands data is maintained outside the database. You will have the same number of elements in this array as you have number of bands.
extent
This is the extent of all the raster rows in your raster set. If you plan to load more data that will change the extent of the set, you'll want to run the DropRasterConstraints function before load and then reapply constraints with AddRasterConstraints after load.
spatial_index
A boolean that is true if raster column has a spatial index.
raster_overviews
catalogs information about raster table columns used for overviews and additional information about them that is useful to know when utilizing overviews. Overview tables are cataloged in both raster_columns
and raster_overviews
because they are rasters in their own right but also serve an additional special purpose of being a lower resolution caricature of a higher resolution table. These are generated along-side the main raster table when you use the -l
switch in raster loading or can be generated manually using AddOverviewConstraints.
Overview tables contain the same constraints as other raster tables as well as additional informational only constraints specific to overviews.
The information in |
Two main reasons for overviews are:
Low resolution representation of the core tables commonly used for fast mapping zoom-out.
Computations are generally faster to do on them than their higher resolution parents because there are fewer records and each pixel covers more territory. Though the computations are not as accurate as the high-res tables they support, they can be sufficient in many rule-of-thumb computations.
The raster_overviews
catalog contains the following columns of information.
o_table_catalog
The database the overview table is in. This will always read the current database.
o_table_schema
The database schema the overview raster table belongs to.
o_table_name
raster overview table name
o_raster_column
the raster column in the overview table.
r_table_catalog
The database the raster table that this overview services is in. This will always read the current database.
r_table_schema
The database schema the raster table that this overview services belongs to.
r_table_name
raster table that this overview services.
r_raster_column
the raster column that this overview column services.
overview_factor
- this is the pyramid level of the overview table. The higher the number the lower the resolution of the table. raster2pgsql if given a folder of images, will compute overview of each image file and load separately. Level 1 is assumed and always the original file. Level 2 is will have each tile represent 4 of the original. So for example if you have a folder of 5000x5000 pixel image files that you chose to chunk 125x125, for each image file your base table will have (5000*5000)/(125*125) records = 1600, your (l=2) o_2
table will have ceiling(1600/Power(2,2)) = 400 rows, your (l=3) o_3
will have ceiling(1600/Power(2,3) ) = 200 rows. If your pixels aren't divisible by the size of your tiles, you'll get some scrap tiles (tiles not completely filled). Note that each overview tile generated by raster2pgsql has the same number of pixels as its parent, but is of a lower resolution where each pixel of it represents (Power(2,overview_factor) pixels of the original).
Il fatto che PostGIS raster vi fornisca le funzioni per restituire i raster in formati di immagine noti vi dà diverse opzioni per visualizzarli. Per esempio potete utilizzare OpenOffice / LibreOffice, come mostrato alla pagina Rendering PostGIS Raster graphics with LibreOffice Base Reports. Inoltre potete utilizzare diversi linguaggi, come illustrato in questo paragrafo.
In questo paragrafo mostreremo come utilizzare il driver PHP PostgreSQL e la famiglia di funzioni ST_AsGDALRaster per estrarre le bande 1,2,3 di un raster a una richiesta PHP che poi può essere inserita in un tag src di un'immagine HTML.
La query di esempio mostra come combinare varie funzioni raster per recuperare tutte le tile che intersecano una data area rettangolare in wgs84, unisce le tile risultanti per tutte le bande con ST_Union, le trasforma in una proiezione specificata dall'utente con ST_Transform e infine crea un PNG in uscita tramite ST_AsPNG.
Andreste a chiamare il codice sotto utilizzando
http://mywebserver/test_raster.php?srid=2249
per ottenere l'immagine raster proiettata nel sistema di riferimento Massachusetts state plane feet.
<?php /** contents of test_raster.php **/ $conn_str ='dbname=mydb host=localhost port=5432 user=myuser password=mypwd'; $dbconn = pg_connect($conn_str); header('Content-Type: image/png'); /**If a particular projection was requested use it otherwise use mass state plane meters **/ if (!empty( $_REQUEST['srid'] ) && is_numeric( $_REQUEST['srid']) ){ $input_srid = intval($_REQUEST['srid']); } else { $input_srid = 26986; } /** The set bytea_output may be needed for PostgreSQL 9.0+, but not for 8.4 **/ $sql = "set bytea_output='escape'; SELECT ST_AsPNG(ST_Transform( ST_AddBand(ST_Union(rast,1), ARRAY[ST_Union(rast,2),ST_Union(rast,3)]) ,$input_srid) ) As new_rast FROM aerials.boston WHERE ST_Intersects(rast, ST_Transform(ST_MakeEnvelope(-71.1217, 42.227, -71.1210, 42.218,4326),26986) )"; $result = pg_query($sql); $row = pg_fetch_row($result); pg_free_result($result); if ($row === false) return; echo pg_unescape_bytea($row[0]); ?>
In questo paragrafo vi mostreremo come usare il driver .NET Npgsql e la famiglia di funzioni ST_AsGDALRaster per inviare in uscita le bande 1,2,3 di un raster a una richiesta PHP che può poi essere inserita nel tag src di un immagine html.
Per questo vi servirà il driver PostgreSQL npgsql .NET. Potete recuperare il più recente da http://npgsql.projects.postgresql.org/. Scaricatelo e salvatelo nella vostra cartella bin di ASP.NET per poter lavorare.
La query di esempio mostra come combinare varie funzioni raster per recuperare tutte le tile che intersecano una data area rettangolare in wgs84, unisce le tile risultanti per tutte le bande con ST_Union, le trasforma in una proiezione specificata dall'utente con ST_Transform e infine crea un PNG in uscita tramite ST_AsPNG.
Questo è lo stesso esempio illustrato in Section 5.3.1, “Esempio di output in PHP, utilizzando ST_AsPNG assieme ad altre funzioni raster”, implementato in C#.
Richiamerete il codice sotto come
http://mywebserver/TestRaster.ashx?srid=2249
per ottenere l'immagine raster nel sistema di riferimento Massachusetts state plane feet.
-- web.config connection string section -- <connectionStrings> <add name="DSN" connectionString="server=localhost;database=mydb;Port=5432;User Id=myuser;password=mypwd"/> </connectionStrings >
// Code for TestRaster.ashx <%@ WebHandler Language="C#" Class="TestRaster" %> using System; using System.Data; using System.Web; using Npgsql; public class TestRaster : IHttpHandler { public void ProcessRequest(HttpContext context) { context.Response.ContentType = "image/png"; context.Response.BinaryWrite(GetResults(context)); } public bool IsReusable { get { return false; } } public byte[] GetResults(HttpContext context) { byte[] result = null; NpgsqlCommand command; string sql = null; int input_srid = 26986; try { using (NpgsqlConnection conn = new NpgsqlConnection(System.Configuration.ConfigurationManager.ConnectionStrings["DSN"].ConnectionString)) { conn.Open(); if (context.Request["srid"] != null) { input_srid = Convert.ToInt32(context.Request["srid"]); } sql = @"SELECT ST_AsPNG( ST_Transform( ST_AddBand( ST_Union(rast,1), ARRAY[ST_Union(rast,2),ST_Union(rast,3)]) ,:input_srid) ) As new_rast FROM aerials.boston WHERE ST_Intersects(rast, ST_Transform(ST_MakeEnvelope(-71.1217, 42.227, -71.1210, 42.218,4326),26986) )"; command = new NpgsqlCommand(sql, conn); command.Parameters.Add(new NpgsqlParameter("input_srid", input_srid)); result = (byte[]) command.ExecuteScalar(); conn.Close(); } } catch (Exception ex) { result = null; context.Response.Write(ex.Message.Trim()); } return result; } }
Questa è una semplice app per console java che prende una query, ne restituisce la corrispondente immagine e la scrive in un file specificato.
Potete scaricare i driver JDBC per PostgreSQL più recenti da http://jdbc.postgresql.org/download.html
Potete compilare il codice seguente con un comando tipo:
set env CLASSPATH .:..\postgresql-9.0-801.jdbc4.jar javac SaveQueryImage.java jar cfm SaveQueryImage.jar Manifest.txt *.class
E lanciarlo dalla riga di comando con un'istruzione tipo
java -jar SaveQueryImage.jar "SELECT ST_AsPNG(ST_AsRaster(ST_Buffer(ST_Point(1,5),10, 'quad_segs=2'),150, 150, '8BUI',100));" "test.png"
-- Manifest.txt -- Class-Path: postgresql-9.0-801.jdbc4.jar Main-Class: SaveQueryImage
// Code for SaveQueryImage.java import java.sql.Connection; import java.sql.SQLException; import java.sql.PreparedStatement; import java.sql.ResultSet; import java.io.*; public class SaveQueryImage { public static void main(String[] argv) { System.out.println("Checking if Driver is registered with DriverManager."); try { //java.sql.DriverManager.registerDriver (new org.postgresql.Driver()); Class.forName("org.postgresql.Driver"); } catch (ClassNotFoundException cnfe) { System.out.println("Couldn't find the driver!"); cnfe.printStackTrace(); System.exit(1); } Connection conn = null; try { conn = DriverManager.getConnection("jdbc:postgresql://localhost:5432/mydb","myuser", "mypwd"); conn.setAutoCommit(false); PreparedStatement sGetImg = conn.prepareStatement(argv[0]); ResultSet rs = sGetImg.executeQuery(); FileOutputStream fout; try { rs.next(); /** Output to file name requested by user **/ fout = new FileOutputStream(new File(argv[1]) ); fout.write(rs.getBytes(1)); fout.close(); } catch(Exception e) { System.out.println("Can't create file"); e.printStackTrace(); } rs.close(); sGetImg.close(); conn.close(); } catch (SQLException se) { System.out.println("Couldn't connect: print out a stack trace and exit."); se.printStackTrace(); System.exit(1); } } }
This is a plpython stored function that creates a file in the server directory for each record. Requires you have plpython installed. Should work fine with both plpythonu and plpython3u.
CREATE OR REPLACE FUNCTION write_file (param_bytes bytea, param_filepath text) RETURNS text AS $$ f = open(param_filepath, 'wb+') f.write(param_bytes) return param_filepath $$ LANGUAGE plpythonu;
--scrive cinque immagini sul server PostgreSQL con dimensioni variabili -- notaer che il l'account da cui gira il demone postgresql deve avere diritto di scrittura sulla cartella -- questo mostra i nomi dei file creati; SELECT write_file(ST_AsPNG( ST_AsRaster(ST_Buffer(ST_Point(1,5),j*5, 'quad_segs=2'),150*j, 150*j, '8BUI',100)), 'C:/temp/slices'|| j || '.png') FROM generate_series(1,5) As j; write_file --------------------- C:/temp/slices1.png C:/temp/slices2.png C:/temp/slices3.png C:/temp/slices4.png C:/temp/slices5.png
Purtroppo PSQL non dispone di una funzione comoda per l'esportazione di file binari. Quello che segue è una soluzione improvvisata derivante dai suggerimenti che trovate alla pagina Clever Trick Challenge -- Outputting bytea with psql e che si appoggia al supporto di PostgreSQL per i tipi "large object". Per utilizzare questo sistema anzitutto lanciate la riga di comando psql e collegatevi al database.
A differenza dell'approcio python, questo sistema crea il file in locale sul vostro computer.
SELECT oid, lowrite(lo_open(oid, 131072), png) As num_bytes FROM ( VALUES (lo_create(0), ST_AsPNG( (SELECT rast FROM aerials.boston WHERE rid=1) ) ) ) As v(oid,png); -- otterrete qualcosa tipo -- oid | num_bytes ---------+----------- 2630819 | 74860 -- prendete nota dell'oid e sostituite c:/test.png con il percorso del file -- sul vostro computer \lo_export 2630819 'C:/temp/aerial_samp.png' -- questo cancella il file dall'archiviazione dei large objects nel database SELECT lo_unlink(2630819);
Table of Contents
The Minnesota MapServer is an internet web-mapping server which conforms to the OpenGIS Web Mapping Server specification.
The MapServer homepage is at http://mapserver.org.
The OpenGIS Web Map Specification is at http://www.opengeospatial.org/standards/wms.
To use PostGIS with MapServer, you will need to know about how to configure MapServer, which is beyond the scope of this documentation. This section will cover specific PostGIS issues and configuration details.
To use PostGIS with MapServer, you will need:
Version 0.6 or newer of PostGIS.
Version 3.5 or newer of MapServer.
MapServer accesses PostGIS/PostgreSQL data like any other PostgreSQL client -- using the libpq
interface. This means that MapServer can be installed on any machine with network access to the PostGIS server, and use PostGIS as a source of data. The faster the connection between the systems, the better.
Compile and install MapServer, with whatever options you desire, including the "--with-postgis" configuration option.
In your MapServer map file, add a PostGIS layer. For example:
LAYER CONNECTIONTYPE postgis NAME "widehighways" # Connect to a remote spatial database CONNECTION "user=dbuser dbname=gisdatabase host=bigserver" PROCESSING "CLOSE_CONNECTION=DEFER" # Get the lines from the 'geom' column of the 'roads' table DATA "geom from roads using srid=4326 using unique gid" STATUS ON TYPE LINE # Of the lines in the extents, only render the wide highways FILTER "type = 'highway' and numlanes >= 4" CLASS # Make the superhighways brighter and 2 pixels wide EXPRESSION ([numlanes] >= 6) STYLE COLOR 255 22 22 WIDTH 2 END END CLASS # All the rest are darker and only 1 pixel wide EXPRESSION ([numlanes] < 6) STYLE COLOR 205 92 82 END END END
In the example above, the PostGIS-specific directives are as follows:
For PostGIS layers, this is always "postgis".
The database connection is governed by the a 'connection string' which is a standard set of keys and values like this (with the default values in <>):
user=<username> password=<password> dbname=<username> hostname=<server> port=<5432>
An empty connection string is still valid, and any of the key/value pairs can be omitted. At a minimum you will generally supply the database name and username to connect with.
The form of this parameter is "<geocolumn> from <tablename> using srid=<srid> using unique <primary key>" where the column is the spatial column to be rendered to the map, the SRID is SRID used by the column and the primary key is the table primary key (or any other uniquely-valued column with an index).
You can omit the "using srid" and "using unique" clauses and MapServer will automatically determine the correct values if possible, but at the cost of running a few extra queries on the server for each map draw.
Putting in a CLOSE_CONNECTION=DEFER if you have multiple layers reuses existing connections instead of closing them. This improves speed. Refer to for MapServer PostGIS Performance Tips for a more detailed explanation.
The filter must be a valid SQL string corresponding to the logic normally following the "WHERE" keyword in a SQL query. So, for example, to render only roads with 6 or more lanes, use a filter of "num_lanes >= 6".
In your spatial database, ensure you have spatial (GiST) indexes built for any the layers you will be drawing.
CREATE INDEX [indexname] ON [tablename] USING GIST ( [geometrycolumn] );
If you will be querying your layers using MapServer you will also need to use the "using unique" clause in your DATA statement.
MapServer requires unique identifiers for each spatial record when doing queries, and the PostGIS module of MapServer uses the unique value you specify in order to provide these unique identifiers. Using the table primary key is the best practice.
The USING
pseudo-SQL clause is used to add some information to help mapserver understand the results of more complex queries. More specifically, when either a view or a subselect is used as the source table (the thing to the right of "FROM" in a DATA
definition) it is more difficult for mapserver to automatically determine a unique identifier for each row and also the SRID for the table. The USING
clause can provide mapserver with these two pieces of information as follows:
DATA "geom FROM ( SELECT table1.geom AS geom, table1.gid AS gid, table2.data AS data FROM table1 LEFT JOIN table2 ON table1.id = table2.id ) AS new_table USING UNIQUE gid USING SRID=4326"
MapServer requires a unique id for each row in order to identify the row when doing map queries. Normally it identifies the primary key from the system tables. However, views and subselects don't automatically have an known unique column. If you want to use MapServer's query functionality, you need to ensure your view or subselect includes a uniquely valued column, and declare it with USING UNIQUE
. For example, you could explicitly select nee of the table's primary key values for this purpose, or any other column which is guaranteed to be unique for the result set.
"Querying a Map" is the action of clicking on a map to ask for information about the map features in that location. Don't confuse "map queries" with the SQL query in a |
PostGIS needs to know which spatial referencing system is being used by the geometries in order to return the correct data back to MapServer. Normally it is possible to find this information in the "geometry_columns" table in the PostGIS database, however, this is not possible for tables which are created on the fly such as subselects and views. So the USING SRID=
option allows the correct SRID to be specified in the DATA
definition.
Lets start with a simple example and work our way up. Consider the following MapServer layer definition:
LAYER CONNECTIONTYPE postgis NAME "roads" CONNECTION "user=theuser password=thepass dbname=thedb host=theserver" DATA "geom from roads" STATUS ON TYPE LINE CLASS STYLE COLOR 0 0 0 END END END
This layer will display all the road geometries in the roads table as black lines.
Now lets say we want to show only the highways until we get zoomed in to at least a 1:100000 scale - the next two layers will achieve this effect:
LAYER CONNECTIONTYPE postgis CONNECTION "user=theuser password=thepass dbname=thedb host=theserver" PROCESSING "CLOSE_CONNECTION=DEFER" DATA "geom from roads" MINSCALE 100000 STATUS ON TYPE LINE FILTER "road_type = 'highway'" CLASS COLOR 0 0 0 END END LAYER CONNECTIONTYPE postgis CONNECTION "user=theuser password=thepass dbname=thedb host=theserver" PROCESSING "CLOSE_CONNECTION=DEFER" DATA "geom from roads" MAXSCALE 100000 STATUS ON TYPE LINE CLASSITEM road_type CLASS EXPRESSION "highway" STYLE WIDTH 2 COLOR 255 0 0 END END CLASS STYLE COLOR 0 0 0 END END END
The first layer is used when the scale is greater than 1:100000, and displays only the roads of type "highway" as black lines. The FILTER
option causes only roads of type "highway" to be displayed.
The second layer is used when the scale is less than 1:100000, and will display highways as double-thick red lines, and other roads as regular black lines.
So, we have done a couple of interesting things using only MapServer functionality, but our DATA
SQL statement has remained simple. Suppose that the name of the road is stored in another table (for whatever reason) and we need to do a join to get it and label our roads.
LAYER CONNECTIONTYPE postgis CONNECTION "user=theuser password=thepass dbname=thedb host=theserver" DATA "geom FROM (SELECT roads.gid AS gid, roads.geom AS geom, road_names.name as name FROM roads LEFT JOIN road_names ON roads.road_name_id = road_names.road_name_id) AS named_roads USING UNIQUE gid USING SRID=4326" MAXSCALE 20000 STATUS ON TYPE ANNOTATION LABELITEM name CLASS LABEL ANGLE auto SIZE 8 COLOR 0 192 0 TYPE truetype FONT arial END END END
This annotation layer adds green labels to all the roads when the scale gets down to 1:20000 or less. It also demonstrates how to use an SQL join in a DATA
definition.
Java clients can access PostGIS "geometry" objects in the PostgreSQL database either directly as text representations or using the JDBC extension objects bundled with PostGIS. In order to use the extension objects, the "postgis.jar" file must be in your CLASSPATH along with the "postgresql.jar" JDBC driver package.
import java.sql.*; import java.util.*; import java.lang.*; import org.postgis.*; public class JavaGIS { public static void main(String[] args) { java.sql.Connection conn; try { /* * Load the JDBC driver and establish a connection. */ Class.forName("org.postgresql.Driver"); String url = "jdbc:postgresql://localhost:5432/database"; conn = DriverManager.getConnection(url, "postgres", ""); /* * Add the geometry types to the connection. Note that you * must cast the connection to the pgsql-specific connection * implementation before calling the addDataType() method. */ ((org.postgresql.PGConnection)conn).addDataType("geometry",Class.forName("org.postgis.PGgeometry")); ((org.postgresql.PGConnection)conn).addDataType("box3d",Class.forName("org.postgis.PGbox3d")); /* * Create a statement and execute a select query. */ Statement s = conn.createStatement(); ResultSet r = s.executeQuery("select geom,id from geomtable"); while( r.next() ) { /* * Retrieve the geometry as an object then cast it to the geometry type. * Print things out. */ PGgeometry geom = (PGgeometry)r.getObject(1); int id = r.getInt(2); System.out.println("Row " + id + ":"); System.out.println(geom.toString()); } s.close(); conn.close(); } catch( Exception e ) { e.printStackTrace(); } } }
The "PGgeometry" object is a wrapper object which contains a specific topological geometry object (subclasses of the abstract class "Geometry") depending on the type: Point, LineString, Polygon, MultiPoint, MultiLineString, MultiPolygon.
PGgeometry geom = (PGgeometry)r.getObject(1); if( geom.getType() == Geometry.POLYGON ) { Polygon pl = (Polygon)geom.getGeometry(); for( int r = 0; r < pl.numRings(); r++) { LinearRing rng = pl.getRing(r); System.out.println("Ring: " + r); for( int p = 0; p < rng.numPoints(); p++ ) { Point pt = rng.getPoint(p); System.out.println("Point: " + p); System.out.println(pt.toString()); } } }
The JavaDoc for the extension objects provides a reference for the various data accessor functions in the geometric objects.
...
Table of Contents
Le attuali versioni di PostgreSQL (compresa la versione 8.0) sono affette da un difetto del query optimizer riguardante le tabelle TOAST. Le tabelle TOAST sono una sorta di "spazio aggiuntivo" usato per memorizzare valori di grandi dimensioni (nel senso di dimensioni dei dati) che non possono essere salvati in normali pagine (per esempio lunghi testi, immagini o geometrie con molti vertici). Vedi la documentazione PostgreSQL sulle tabelle TOAST per maggiori informazioni).
Il problema si presenta nel caso di tabelle contenenti geometrie di grandi dimensioni ma poche righe (per esempio una tabella contenente i confini di tutti gli stati europei ad alta risoluzione). In questo caso infatti la tabella in se è di piccole dimensioni ma usa molto spazio TOAST. Nel nostro esempio la tabella conteneva circa 80 righe e utilizzava solo 3 pagine di dati, ma la tabella TOAST ne utilizzava 8225.
Ora si lanci una query che usi l'operatore && e che [MATCHES] solo poche righe. Il
Per controllare se i propri dati sono interessati da questo bug, si può usare il comando PostgreSQL "EXPLAIN ANALYZE". Per maggiori informazioni e dettagli tecnici consultare il corrispondente thread sulla mailing list di PostgreSQL: http://archives.postgresql.org/pgsql-performance/2005-02/msg00030.php
Gli sviluppatori di PostgreSQL stanno cercando di risolvere il problema rendendo la valutazione della query indipendente dalla tabella TOAST. Per ora ci sono due possibili soluzioni alternative:
La è forzare il query planner ad usare l'indice spaziale usando il comando "SET enable_seqscan TO off;" prima di lanciare la query. Questo comando impedisce al query planner di usare lo scan sequenziale della tabella se possibile e lo forza quindi ad usare l'indice GIST. Tuttavia il comando deve essere lanciato ad ogni connessione e, per evitare di confondere il query planner in altri casi, il parametro deve essere resettato dopo l'esecuzione della query interessata con il comando "SET enable_seqscan TO on;" .
Il secondo metodo è rendere lo scan sequenziale così veloce come il query planner si aspetta che sia. Questo può essere raggiunto aggiungendo una colonna addizionale in cui salvare la bounding box di ogni geometria. Nel nostro esempio i comandi sarebbero:
SELECT AddGeometryColumn('myschema','mytable','bbox','4326','GEOMETRY','2'); UPDATE mytable SET bbox = ST_Envelope(ST_Force_2d(the_geom));
Ora la query deve essere modificata in modo da usare l'operatore && con la colonna bbox piuttosto che con la colonna geom_column:
SELECT geom_column FROM mytable WHERE bbox && ST_SetSRID('BOX3D(0 0,1 1)'::box3d,4326);
Ovviamente la colonna bbox deve essere mantenuta attuale quando si modificano o si aggiungono geometrie. La via più semplice per fare questo sarebbe con un trigger, oppure l'applicazione può essere modificata in modo da attualizzare anche la colonna bbox oppure si può lanciare l'UPDATE precedente dopo ogni modifica.
Per tabelle che vengono per lo più solo lette, e dove un singolo indice è usato dalla maggior parte delle query, PostgreSQL offre il comando CLUSTER. Questo comando riordina fisicamente le righe in modo che l'ordine corrisponda a quello dell'indice. Con questo metodo si migliorano le prestazioni per due motivi: primo, il numero delle ricerche nella tabella dei dati è ridotto drasticamente. Secondo, se i dati interessati dalla query sono concentrati in un piccolo intervallo sull'indice, il processo di mettere in cache sarà più efficiente perché le righe saranno distribuite all'interno di poche pagine. (si invita a leggere la documentazione di PostgreSQL riguardante il comando CLUSTER).
Attualmente però, PostgreSQL non permette di usare il clustering con indici GIST, perché gli indici GIST ignorano i valori nulli:
lwgeom=# CLUSTER my_geom_index ON my_table; ERROR: cannot cluster when index access method does not handle null values HINT: You may be able to work around this by marking column "the_geom" NOT NULL.
Come suggerito dal messaggio di errore, è possibile aggirare il problema aggiungendo un vincolo "NOT NULL" alla tabella:
lwgeom=# ALTER TABLE my_table ALTER COLUMN the_geom SET not null; ALTER TABLE
Ovviamente questo non funzionerà se la colonna the_geom già contiene valori nulli. Inoltre il vincolo dev'essere definito usando il comando precedente. Usare un vincolo CHECK del tipo "ALTER TABLE blubb ADD CHECK (geometry is not null);" non funzionerà.
A volte può accadere di avere dati in 3D o 4D, ma di accederli sempre usando funzioni che danno in output solo geometrie 2D come ST_AsText() oppure ST_AsBinary(). Queste funzioni processano internamente le geometrie eseguendo ST_Force2D() e questo può risultare in un peggioramento delle prestazioni significativo nel caso di geometrie di grandi dimensioni. Per evitare questo problema è consigliabile eliminare le dimensioni non utilizzate una volte e per sempre:
UPDATE mytable SET the_geom = ST_Force_2d(the_geom); VACUUM FULL ANALYZE mytable;
Se la colonna di tipo geometry è stata aggiunta usando la funzione AddGeometryColumn(), verrà creato anche un vincolo dimensionale sulla geometria. Per aggirare il vincolo sarà necessario cancellarlo. Ricorda di attualizzare il record nella tabella geometry_columns e di ricreare il vincolo successivamente.
Nel caso di grandi tabelle può essere sensato dividere l'UPDATE in porzioni più piccole, usando la clausola WHERE con la chiave primaria o un altro criterio flessibile per attualizzare solo una parte della tabella , ed eseguendo un semplice "VACUUM;" tra gli UPDATE. Questo accorgimento ridurrà drasticamente il bisogno lo spazio temporaneo sul disco. Inoltre, se la colonna geometrica contiene geometrie con diverse dimensioni, si può limitare l'UPDATE con "WHERE dimension(the_geom)>2", evitando così di riscrivere le geometrie che sono già in 2D.
Questi consigli sono tratti dalla presentazione di Kevin Neufeld "Tips for the PostGIS Power User" alla conferenza FOSS4G 2007. A seconda del contesto in cui viene usato PostGIS (per esempio dati statici e analisi complesse piuttosto che update frequenti e numerosi utenti) questi cambiamenti possono significare un significativo aumento della velocità delle query.
Per più consigli (e una formattazione migliore) la presentazione originale è consultabile a http://2007.foss4g.org/presentations/view.php?abstract_id=117.
Queste impostazioni sono definite in postgresql.conf:
Numero massimo di segmenti del log file tra checkpoint WAL (ogni segmento è normalmente 16MB); l'impostazione predefinita è 3.
Impostare almeno su 10 o 30 per database con intensa attività di scrittura, o più quando si importa nel database grandi quantità di dati. Un altro articolo da leggere sull'argomento è Greg Smith: Checkpoint and Background writer
Se possibile, salvare il xlog su un disco separato
Il valore di default è "off" per versioni inferiori alla 8.4 e "partition" a partire dalla 8.4()
Generalmente usato per il partizionamento delle tabelle. Con versioni di PostgreSQL inferiori alla 8.4 impostare il parametro su "on" per permettere al query planner di ottimizzare secondo necessità. Da PostgreSQL 8.4, il valore di default è "partition", ideale per versioni 8.4 e superiori perché forza il query planner ad analizzare i vincoli delle tabelle solo se queste sono in una gerarchia di ereditarietà e non essere penalizzate dal planner in caso contrario.
Default: ~32MB
Impostare da 1/3 a 3/4 della RAM disponibile.
work_mem (memoria usata per operazioni di ordinamento e query complesse)
Default: 1MB
Aumentare il valore per grandi database, query complesse, abbondanza di RAM
Diminuire il valore nel caso di molti utenti concorrenti oppure disponibilità di RAM ridotta
Nel caso di molta RAM e pochi programmatori:
SET work_mem TO 1200000;
maintenance_work_mem (usato per VACUUM, CREATE INDEX, etc.)
Default: 16MB
Generalmente troppo basso. Tiene occupati i processi Input/Output e blocca oggetti quando viene impegnata la memoria swap.
Si raccomandano 32MB oppure 256MB su server di produzione con molta RAM, ma dipende dal numero di utenti concorrenti. Nel caso di molta RAM e pochi sviluppatori:
SET maintainence_work_mem TO 1200000;
Table of Contents
Le funzioni elencate sotto sono quelle di cui un utente PostGIS avrà più probabilmente bisogno. Ci sono altre funzioni, di supporto agli oggetti PostGIS, che non sono utili all'utente comune.
PostGIS ha iniziato una transizione dalla namin convention esistente a una convenzione SQL-MM-centrica. Di conseguenza, molte funzioni di uso comune sono state rinominate usando il prefisso standard "spatial type" (ST). Le funzioni precedenti sono tuttora disponibili, anche se non elencate nel presente documento. Al loro posto sono presenti le funzioni aggiornate corrispondenti. Le funzioni non ST_ che mancano in questo documento sono deprecate e verranno eliminate in una futura release, quindi NON VANNO PIÙ USATE. |
Abstract
Questa sezione illustra i tipi di dati di PostgreSQL che vengono installati da Postgis. Vengono illustrati anche i metodi per eseguire il cast tra diversi tipi, cosa particolarmente importante per la realizzazione di funzioni proprie.
Effettuare il cast significa forzare un tipo in un altro tipo di dato. PostgreSQL è tra i pochi database che offrono la possibilità di definire comportamento di casting per tipi personalizzati e le funzioni usate per il casting. Un cast può essere definito come automatico: non è necessario utilizzare la funzione CAST(myfoo As otherfootype) oppure myfoo::otherfootype se si esegue una funzione che supporta unicamente il tipo otherfootype. In questo caso il cast avverrà automaticamente.
Un rischio dell'usare un cast automatico può esserci quando si ha una funzione sovraccaricata, per esempio una funzione che accetta come argomento un box2d e un'altra che accetta un box3d ma nessuna che accetti un geometry. In questo caso le due funzioni possono essere utilizzate con un tipo geometry, perché geometry supporta il cast automatico verso box2d e box3d, e si avrà quindi un errore che indica che la funzione è ambigua. Per forzare PostgreSQL ad effettuare la scelta giusta bisogna effettuare un cast specifico con CAST(mygeom AS box3d) oppure mygeom::box3d
A partire dalla versione PostgreSQL 8.3 - È possibile effettuare il CAST di ogni tipo a "text" (presumibilmente per merito del magico tipo "unknown"), quindi non è necessario definire esplicitamente il CAST di un oggetto a "text".
box2d — Un rettangolo composto dalle coordinate xmin,ymin,xmax,ymax. Spesso è usato per ritornare l'estensione 2D della geometria.
box3d — Un parallelepipedo composto dalle coordinate xmin, ymin, zmin, xmax, ymax, zmax. Spesso usato per ritornare l'estensione 3D di una geometria o di una collezione di geometrie.
geometry — Tipi di dati spaziali piani.
Il tipo di dato "geometry" è un fondamentale tipo spaziale di Postgis usato per rappresentare una entità in un sistema di coordinate euclideo.
geometry_dump — Un tipo di dato spaziale composto da due campi: geom (contenente un oggetto geometry) e path[] (un array unidimensionale 1-d contenente la posizione della geometria all'interno della collezione).
geography — Tipi spaziali ellissoidi.
geography è un tipo spaziale usato per rappresentare un'entità in un sistema di coordinate sferico valido per l'intero pianeta.
postgis.backend — The backend to service a function where GEOS and SFCGAL overlap. Options: geos or sfcgal. Defaults to geos.
This GUC is only relevant if you compiled PostGIS with sfcgal support. By default geos
backend is used for functions where both GEOS and SFCGAL have the same named function. This variable allows you to override and make sfcgal the backend to service the request.
Availability: 2.1.0
postgis.gdal_datapath — A configuration option to assign the value of GDAL's GDAL_DATA option. If not set, the environmentally set GDAL_DATA variable is used.
A PostgreSQL GUC variable for setting the value of GDAL's GDAL_DATA option. The postgis.gdal_datapath
value should be the complete physical path to GDAL's data files.
This configuration option is of most use for Windows platforms where GDAL's data files path is not hard-coded. This option should also be set when GDAL's data files are not located in GDAL's expected path.
This option can be set in PostgreSQL's configuration file postgresql.conf. It can also be set by connection or transaction. |
Availability: 2.2.0
Additional information about GDAL_DATA is available at GDAL's Configuration Options. |
postgis.gdal_enabled_drivers — A configuration option to set the enabled GDAL drivers in the PostGIS environment. Affects the GDAL configuration variable GDAL_SKIP.
A configuration option to set the enabled GDAL drivers in the PostGIS environment. Affects the GDAL configuration variable GDAL_SKIP. This option can be set in PostgreSQL's configuration file: postgresql.conf. It can also be set by connection or transaction.
The initial value of postgis.gdal_enabled_drivers
may also be set by passing the environment variable POSTGIS_GDAL_ENABLED_DRIVERS
with the list of enabled drivers to the process starting PostgreSQL.
Enabled GDAL specified drivers can be specified by the driver's short-name or code. Driver short-names or codes can be found at GDAL Raster Formats. Multiple drivers can be specified by putting a space between each driver.
There are three special codes available for
When |
In the standard PostGIS installation, |
Additional information about GDAL_SKIP is available at GDAL's Configuration Options. |
Availability: 2.2.0
Set and reset postgis.gdal_enabled_drivers
Sets backend for all new connections to database
ALTER DATABASE mygisdb SET postgis.gdal_enabled_drivers TO 'GTiff PNG JPEG';
Sets default enabled drivers for all new connections to server. Requires super user access and PostgreSQL 9.4+. Also not that database, session, and user settings override this.
ALTER SYSTEM SET postgis.gdal_enabled_drivers TO 'GTiff PNG JPEG'; SELECT pg_reload_conf();
SET postgis.gdal_enabled_drivers TO 'GTiff PNG JPEG'; SET postgis.gdal_enabled_drivers = default;
Enable all GDAL Drivers
SET postgis.gdal_enabled_drivers = 'ENABLE_ALL';
Disable all GDAL Drivers
SET postgis.gdal_enabled_drivers = 'DISABLE_ALL';
postgis.enable_outdb_rasters — A boolean configuration option to enable access to out-db raster bands.
A boolean configuration option to enable access to out-db raster bands. This option can be set in PostgreSQL's configuration file: postgresql.conf. It can also be set by connection or transaction.
The initial value of postgis.enable_outdb_rasters
may also be set by passing the environment variable POSTGIS_ENABLE_OUTDB_RASTERS
with a non-zero value to the process starting PostgreSQL.
Even if |
In the standard PostGIS installation, |
Availability: 2.2.0
geometry_columns
. Per default la funzione converte tutte le colonne geometry definite senza typemod in geometry con typemod. Per ottenere il vecchio meccanismo con i vincoli spaziali si può settare la variabile use_typmod=false
AddGeometryColumn — Aggiunge una colonna di tipo geometry in una tabella già esistente. Per default usa il modificatore di tipo per la definizione della geometria piuttosto che dei vincoli. Si può passare alla funzione il parametro use_typemod per attivare il vecchio meccanismo basato sui vincoli.
text AddGeometryColumn(
varchar table_name, varchar column_name, integer srid, varchar type, integer dimension, boolean use_typmod=true)
;
text AddGeometryColumn(
varchar schema_name, varchar table_name, varchar column_name, integer srid, varchar type, integer dimension, boolean use_typmod=true)
;
text AddGeometryColumn(
varchar catalog_name, varchar schema_name, varchar table_name, varchar column_name, integer srid, varchar type, integer dimension, boolean use_typmod=true)
;
Aggiunge una colonna di tipo geometry ad una tabella già esistente. schema_name
è il nome dello schema contenente la tabella. srid
deve essere un numero intero che si riferisce a un record presente nella tabella SPATIAL_REF_SYS. type
deve essere una stringa corrispondente al tipo di geometria, per esempio 'POLYGON' oppure 'MULTILINESTRING'. La funzione produce un errore se lo schema non esiste (oppure non è visibile nel search_path attual), o la SRID specificata, il tipo di geometria o la dimensione sono invalidi.
Cambiamento nella versione 2.0.0: questa funzione non aggiorna più geometry_columns perché geometry_columns non è più una tabella ma una vista che estrae automaticamente le informazioni necessarie dal system catalog. Inoltre per default la funzione non crea vincoli ma usa il modificatore di tipi integrato in PostgreSQL. Per esempio: creare una colonna con tipo POINT e con SRID 4326 con questa funzione, ora è equivalente a: Cambiamento nella versione 2.0.0: il vecchio funzionamento con i vincoli può essere attivato passando alla funzione l'argomento |
Cambiamento in versione 2.0.0: le viste non possono più essere registrate in geometry_columns, a meno che le colonne geometry a cui fanno riferimento non siano state generate con typmod e usate senza funzioni wrapper. In questo caso la vista sarà registrata correttamente in geometry_columns perché eredita il typmod dalla colonna geometry originale. Le viste che usano funzioni che ritornano geometrie verranno registrate correttamente se il tipo delll'output della funzione verrà definito esplicitamente con la funzione CAST come typmod geometry. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
Miglioramento nella version 2.0.0: introdotto il parametro use_typmod. Se settato su true (o se omesso) la funzione genererà una colonna geometry basata su typmod. Se settato su "false", la funzione genererà una colonna geometry con basata su vincoli geometrici.
-- Creiamo uno schema in sui salvare i dati CREATE SCHEMA my_schema; -- Creiamo una Semplice tabella CREATE TABLE my_schema.my_spatial_table (id serial); -- La descrizione mostra una semplice tabella con un'unica colonna "id" postgis=# \d my_schema.my_spatial_table Table "my_schema.my_spatial_table" Column | Type | Modifiers --------+---------+------------------------------------------------------------------------- id | integer | not null default nextval('my_schema.my_spatial_table_id_seq'::regclass) -- Aggiungiamo una colonna geometry SELECT AddGeometryColumn ('my_schema','my_spatial_table','geom',4326,'POINT',2); -- Aggiungiamo un'altra colonna geometry settando il parametro use_typmod = false, -- generando quindo una colonna geometry basata su vincoli SELECT AddGeometryColumn ('my_schema','my_spatial_table','geom_c',4326,'POINT',2, false); --Di nuovo aggiungiamo un'altra colonna geometry usando il vecchio meccanismo con vincoli Add a curvepolygon using old constraint behavior SELECT AddGeometryColumn ('my_schema','my_spatial_table','geomcp_c',4326,'CURVEPOLYGON',2, false); -- Descriviamo di nuovo la tabella con le nuove 3 colonne geometry \d my_schema.my_spatial_table addgeometrycolumn ------------------------------------------------------------------------- my_schema.my_spatial_table.geomcp_c SRID:4326 TYPE:CURVEPOLYGON DIMS:2 (1 row) Table "my_schema.my_spatial_table" Column | Type | Modifiers ----------+----------------------+------------------------------------------------------------------------- id | integer | not null default nextval('my_schema.my_spatial_table_id_seq'::regclass) geom | geometry(Point,4326) | geom_c | geometry | geomcp_c | geometry | Check constraints: "enforce_dims_geom_c" CHECK (st_ndims(geom_c) = 2) "enforce_dims_geomcp_c" CHECK (st_ndims(geomcp_c) = 2) "enforce_geotype_geom_c" CHECK (geometrytype(geom_c) = 'POINT'::text OR geom_c IS NULL) "enforce_geotype_geomcp_c" CHECK (geometrytype(geomcp_c) = 'CURVEPOLYGON'::text OR geomcp_c IS NULL) "enforce_srid_geom_c" CHECK (st_srid(geom_c) = 4326) "enforce_srid_geomcp_c" CHECK (st_srid(geomcp_c) = 4326) -- Le colonne geometry sono automaticamante registrate in geometry_columns SELECT f_geometry_column As col_name, type, srid, coord_dimension As ndims FROM geometry_columns WHERE f_table_name = 'my_spatial_table' AND f_table_schema = 'my_schema'; col_name | type | srid | ndims ----------+--------------+------+------- geom | Point | 4326 | 2 geom_c | Point | 4326 | 2 geomcp_c | CurvePolygon | 4326 | 2
DropGeometryColumn — Rimuove una colonna geometry da una tabella spaziale
text DropGeometryColumn(
varchar table_name, varchar column_name)
;
text DropGeometryColumn(
varchar schema_name, varchar table_name, varchar column_name)
;
text DropGeometryColumn(
varchar catalog_name, varchar schema_name, varchar table_name, varchar column_name)
;
Rimuove una colonna geometry da una tabella spaziale. Il campo schema_name deve corrispondere al campo f_table_schema in geometry_columns.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
Cambiamento nella versione 2.0.0: questa funzione è mantenuta per retrocompatibilità. Attualmente, essendo geometry_columns una vista basata sul system catalog, una colonna geometry può essere rimossa come qualsiasi altra colonna con |
SELECT DropGeometryColumn ('my_schema','my_spatial_table','geom'); ----RESULT output --- dropgeometrycolumn ------------------------------------------------------ my_schema.my_spatial_table.geom effectively removed. -- In PostGIS 2.0+ questo comando è equivalente al comando standard -- alter table. Ambedue i comandi rimuoveranno la colonna da geometry_columns ALTER TABLE my_schema.my_spatial_table DROP column geom;
DropGeometryTable — Rimuove una tabella e tutte le sue referenze da geometry_columns
boolean DropGeometryTable(
varchar table_name)
;
boolean DropGeometryTable(
varchar schema_name, varchar table_name)
;
boolean DropGeometryTable(
varchar catalog_name, varchar schema_name, varchar table_name)
;
Rimuove una tabella spaziale e tutte le sue referenze da geometry_columns. Nota: utilizza la funzione PostgreSQL current_schema() se lo schema non è passato come argomento.
Cambiamento nella versione 2.0.0: questa funzione è mantenuta per retrocompatibilità. Attualmente, essendo geometry_columns una vista basata sul system catalog, una tabella spaziale può essere rimossa come qualsiasi altra tabella con |
PostGIS_Full_Version — Riporta informazioni sulla versione completa di Postgis e sulla configurazione della compilazione.
text PostGIS_Full_Version(
)
;
Riporta informazioni sulla versione completa di Postgis e sulla configurazione della compilazione. Effettua anche un confronto tra librerie e script e suggerisce l'aggiornamento dove ce ne sia bisogno.
SELECT PostGIS_Full_Version(); postgis_full_version ---------------------------------------------------------------------------------- POSTGIS="2.2.0dev r12699" GEOS="3.5.0dev-CAPI-1.9.0 r3989" SFCGAL="1.0.4" PROJ="Rel. 4.8.0, 6 March 2012" GDAL="GDAL 1.11.0, released 2014/04/16" LIBXML="2.7.8" LIBJSON="0.12" RASTER (1 row)
PostGIS_GEOS_Version — Riporta il numero della versione della libreria GEOS.
text PostGIS_GEOS_Version(
)
;
Riporta il numero di versione della libreria GEOS oppure NULL
se il supporto a GEOS non è attivato.
PostGIS_LibXML_Version — Riporta il numero della versione della libreria libxml2.
text PostGIS_LibXML_Version(
)
;
PostGIS_Lib_Build_Date — Riporta la data della compilazione della libreria PostGIS.
text PostGIS_Lib_Build_Date(
)
;
PostGIS_Lib_Version — Riporta il numero di versione della libreria PostGIS.
text PostGIS_Lib_Version(
)
;
PostGIS_PROJ_Version — Riporta il numero di versione della libreria PROJ4.
text PostGIS_PROJ_Version(
)
;
Riporta il numero di versione della libreria PROJ4, oppure NULL
se il supporto a PROJ4 non è attivato.
PostGIS_Scripts_Build_Date — Riporta la data della compilazione degli script PostGIS.
text PostGIS_Scripts_Build_Date(
)
;
Riporta la data della compilazione degli script PostGIS.
Disponibilità: dalla versione 1.0.0RC1
PostGIS_Scripts_Installed — Riporta il numero della versione degli script PostGIS installati del database.
text PostGIS_Scripts_Installed(
)
;
Riporta il numero della versione degli script PostGIS installati del database.
Se l'output di questa funzione non corrisponde all'output di PostGIS_Scripts_Released, probabilmente gli script installati nel database devono essere aggiornati. Vedi la sezione Aggiornamento per più informazioni. |
Disponibilità: dalla versione 0.9.0
PostGIS_Scripts_Released — Riporta il numero di versione dello script postgis.sql rilasciato con la libreria PostGIS installata.
text PostGIS_Scripts_Released(
)
;
Riporta il numero di versione dello script postgis.sql rilasciato con la libreria PostGIS installata.
Dalla versione 1.1.0 questa funzione ritorna lo stesso valore di PostGIS_Lib_Version. Viene mantenuta per retrocompatibilità. |
Disponibilità: dalla versione 0.9.0
PostGIS_Version — Riporta il numero della versione di PostGIS e le opzioni della compilazione.
text PostGIS_Version(
)
;
Populate_Geometry_Columns — Si accerta che le colonne geometry siano definite con typemod oppure abbiano gli appropriati vincoli spaziali e conseguentemente siano correttamente registrate nella vista geometry_columns
. Per default la funzione converte tutte le colonne geometry definite senza typemod in geometry con typemod. Per ottenere il vecchio meccanismo con i vincoli spaziali si può settare la variabile use_typmod=false
text Populate_Geometry_Columns(
boolean use_typmod=true)
;
int Populate_Geometry_Columns(
oid relation_oid, boolean use_typmod=true)
;
Si accerta che le colonne geometry siano definite con typemod oppure abbiano gli appropriati vincoli spaziali e conseguentemente siano correttamente registrate nella vista geometry_columns
.
For backwards compatibility and for spatial needs such as table inheritance where each child table may have different geometry type, the old check constraint behavior is still supported. If you need the old behavior, you need to pass in the new optional argument as false use_typmod=false
. When this is done geometry columns will be created with no type modifiers but will have 3 constraints defined. In particular, this means that every geometry column belonging to a table has at least three constraints:
enforce_dims_the_geom
- assicura che tutte le geometrie hanno la stessa dimensione (vedi ST_NDims)
enforce_geotype_the_geom
- assicura che tutte le geometrie sono dello stesso tipo (vedi GeometryType)
enforce_srid_the_geom
- - assicura che tutte le geometrie hanno la stessa proiezione (vedi ST_SRID)
Se alla funzione viene passato l' oid
di una tabella, questa cerca di determinare srid, dimensione e tipo della geometria di tutte le colonne geometry della tabella, aggiungendo i vincoli se necessario. In caso di successo, una riga viene inserita nella tabella geometry_columns, altrimenti viene lanciata un'eccezione con un messaggio che descrive il problema.
Se alla funzione viene passato l'oid
di una vista, questa cerca di determinare srid, dimensione e tipo della geometria di tutte le colonne geometry della vista, inserendo le appropriate righe nella tabella geometry_columns. I vincoli spaziali non vengono controllati né definiti.
La variante senza parametri è un semplice wrapper per la variante parametrizzata, che prima svuota e poi riempie la tabella geometry_columns per ogni vista e tabella spaziale nel database, aggiungendo gli appropriate vincoli spaziali. La funzione ritorna un sommario con il numero delle colonne geometry trovate e di quelle inserite in geometry_columns
. La versione parametrizzata ritorna semplicemente il numero di righe inserite nella tabella geometry_columns
.
Disponibilità: 1.4.0
Cambiamento nella versione 2.0.0: Per default la funzione utilizza ora type modifier invece di controllare i vincoli spaziali. il meccanismo con i vincoli può essere comunque attivato passando il parametro use_typmod
e settandolo come false.
Miglioramento nelle version 2.0.0: Il parametro opzionale use_typmod
è stato introdotto per permettere di controllare se le colonne devono essere create con typmodifier oppure con i vincoli spaziali.
CREATE TABLE public.myspatial_table(gid serial, geom geometry); INSERT INTO myspatial_table(geom) VALUES(ST_GeomFromText('LINESTRING(1 2, 3 4)',4326) ); -- Il comando seguente usa typ modifiers. Per poter funzionare devono essere presenti dati nella tabella. SELECT Populate_Geometry_Columns('public.myspatial_table'::regclass); populate_geometry_columns -------------------------- 1 \d myspatial_table Table "public.myspatial_table" Column | Type | Modifiers --------+---------------------------+--------------------------------------------------------------- gid | integer | not null default nextval('myspatial_table_gid_seq'::regclass) geom | geometry(LineString,4326) |
-- Il comando seguente modificherà le colonne geometry usando i vincoli spaziali se non sono state definite con typmod oppure non hanno già i vincoli spaziali. --Per poter funzionare devono essere presenti dati nella tabella. CREATE TABLE public.myspatial_table_cs(gid serial, geom geometry); INSERT INTO myspatial_table_cs(geom) VALUES(ST_GeomFromText('LINESTRING(1 2, 3 4)',4326) ); SELECT Populate_Geometry_Columns('public.myspatial_table_cs'::regclass, false); populate_geometry_columns -------------------------- 1 \d myspatial_table_cs Table "public.myspatial_table_cs" Column | Type | Modifiers --------+----------+------------------------------------------------------------------ gid | integer | not null default nextval('myspatial_table_cs_gid_seq'::regclass) geom | geometry | Check constraints: "enforce_dims_geom" CHECK (st_ndims(geom) = 2) "enforce_geotype_geom" CHECK (geometrytype(geom) = 'LINESTRING'::text OR geom IS NULL) "enforce_srid_geom" CHECK (st_srid(geom) = 4326)
UpdateGeometrySRID — Updates the SRID of all features in a geometry column, geometry_columns metadata and srid. If it was enforced with constraints, the constraints will be updated with new srid constraint. If the old was enforced by type definition, the type definition will be changed.
text UpdateGeometrySRID(
varchar table_name, varchar column_name, integer srid)
;
text UpdateGeometrySRID(
varchar schema_name, varchar table_name, varchar column_name, integer srid)
;
text UpdateGeometrySRID(
varchar catalog_name, varchar schema_name, varchar table_name, varchar column_name, integer srid)
;
Aggiorna lo SRID di tutti i record in una colonna geometry, aggiornando anche geometry_columns e il vincolo SRID della colonna. Usa la funzione current_schema() se lo schema non è passato come argomento.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This will change the srid of the roads table to 4326 from whatever it was before
SELECT UpdateGeometrySRID('roads','geom',4326);
The prior example is equivalent to this DDL statement
ALTER TABLE roads ALTER COLUMN geom TYPE geometry(MULTILINESTRING, 4326) USING ST_SetSRID(geom,4326);
If you got the projection wrong (or brought it in as unknown) in load and you wanted to transform to web mercator all in one shot You can do this with DDL but there is no equivalent PostGIS management function to do so in one go.
ALTER TABLE roads ALTER COLUMN geom TYPE geometry(MULTILINESTRING, 3857) USING ST_Transform(ST_SetSRID(geom,4326),3857) ;
LINESTRING
from WKB with the given SRIDST_BdPolyFromText — Costruisce un poligono an partire da una collezione arbitraria di linee chiuse in forma di multilinee rappresentate come testo Well-Known
geometry ST_BdPolyFromText(
text WKT, integer srid)
;
Costruisce un poligono an partire da una collezione arbitraria di linee chiuse in forma di multilinee rappresentate come testo Well-Known
Ritorna un errore se il testo WKT non è una MULTILINESTRING oppure se l'output è un MULTIPOLYGON; in questo caso si può usare ST_BdMPolyFromText oppure ST_BuildArea() per un approccio specifica postgis. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
Disponibilità: 1.1.0 - richiede GEOS >= 2.1.0.
ST_BdMPolyFromText — Costruisce un MultiPolygon a partire da una collezione arbitraria di linee chiuse sotto forma di MultiLineString in formato Well-Known-Text.
geometry ST_BdMPolyFromText(
text WKT, integer srid)
;
Costruisse un poligono a partire da una collezione arbitraria di linee chiuse, poligoni e MultiLineString in formato Well-Known-Text.
Ritorna un errore se il WKT non è una MULTILINESTRING. L'output è MULTIPOLYGON anche se il risultato è un singolo poligono. Usare ST_BdPolyFromText se si è sicuri che il risultato è un singolo poligono oppure vedere ST_BuildArea() per un approccio specifico postgis. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
Disponibilità: 1.1.0 - richiede GEOS >= 2.1.0.
ST_Box2dFromGeoHash — Return a BOX2D from a GeoHash string.
box2d ST_Box2dFromGeoHash(
text geohash, integer precision=full_precision_of_geohash)
;
Return a BOX2D from a GeoHash string.
If no precision
is specficified ST_Box2dFromGeoHash returns a BOX2D based on full precision of the input GeoHash string.
If precision
is specified ST_Box2dFromGeoHash will use that many characters from the GeoHash to create the BOX2D. Lower precision values results in larger BOX2Ds and larger values increase the precision.
Availability: 2.1.0
SELECT ST_Box2dFromGeoHash('9qqj7nmxncgyy4d0dbxqz0'); st_geomfromgeohash -------------------------------------------------- BOX(-115.172816 36.114646,-115.172816 36.114646) SELECT ST_Box2dFromGeoHash('9qqj7nmxncgyy4d0dbxqz0', 0); st_box2dfromgeohash ---------------------- BOX(-180 -90,180 90) SELECT ST_Box2dFromGeoHash('9qqj7nmxncgyy4d0dbxqz0', 10); st_box2dfromgeohash --------------------------------------------------------------------------- BOX(-115.17282128334 36.1146408319473,-115.172810554504 36.1146461963654)
ST_GeogFromText — Ritorna un valore geography sotto forma di Well-Know-Text (WKT) oppure di Extended-Well-Know-Text (EWKT)
geography ST_GeogFromText(
text EWKT)
;
Returns a geography object from the well-known text or extended well-known representation. SRID 4326 is assumed if unspecified. This is an alias for ST_GeographyFromText. Points are always expressed in long lat form.
--- converting lon lat coords to geography ALTER TABLE sometable ADD COLUMN geog geography(POINT,4326); UPDATE sometable SET geog = ST_GeogFromText('SRID=4326;POINT(' || lon || ' ' || lat || ')'); --- specify a geography point using EPSG:4267, NAD27 SELECT ST_AsEWKT(ST_GeogFromText('SRID=4267;POINT(-77.0092 38.889588)'));
ST_GeographyFromText — Ritorna un valore geography sotto forma di Well-Know-Text (WKT) oppure di Extended-Well-Know-Text (EWKT)
geography ST_GeographyFromText(
text EWKT)
;
ST_GeogFromWKB — Crea un oggetto geography a partire da una geometria in Well-Known Binary (WKB) oppure Extended Well-Known Binary (EWKB).
geography ST_GeogFromWKB(
bytea wkb)
;
La funzione ST_GeogFromWKB
accetta come argomento una geometria POstGIS in formato Well-Known-Binary (WKB) oppure Extended WKB a crea un oggetto dell'appropriato tipo geography. Questa funzione ha un ruolo nella Geometry Factory in SQL.
Se non specificato, lo SRID di default è 4326 (WGS 84 long lat).
This method supports Circular Strings and Curves
--Anche se bytea rep contiene singoli \, these need to be escaped when inserting into a table SELECT ST_AsText( ST_GeogFromWKB(E'\\001\\002\\000\\000\\000\\002\\000\\000\\000\\037\\205\\353Q\\270~\\\\\\300\\323Mb\\020X\\231C@\\020X9\\264\\310~\\\\\\300)\\\\\\217\\302\\365\\230C@') ); st_astext ------------------------------------------------------ LINESTRING(-113.98 39.198,-113.981 39.195) (1 row)
ST_GeomFromTWKB — Creates a geometry instance from a TWKB ("Tiny Well-Known Binary") geometry representation.
geometry ST_GeomFromTWKB(
bytea twkb)
;
The ST_GeomFromTWKB
function, takes a a TWKB ("Tiny Well-Known Binary") geometry representation (WKB) and creates an instance of the appropriate geometry type.
SELECT ST_AsText(ST_GeomFromTWKB(ST_AsTWKB('LINESTRING(126 34, 127 35)'::geometry))); st_astext ----------------------------- LINESTRING(126 34, 127 35) (1 row) SELECT ST_AsEWKT( ST_GeomFromTWKB(E'\\x620002f7f40dbce4040105') ); st_asewkt ------------------------------------------------------ LINESTRING(-113.98 39.198,-113.981 39.195) (1 row)
ST_GeomCollFromText — Makes a collection Geometry from collection WKT with the given SRID. If SRID is not give, it defaults to 0.
geometry ST_GeomCollFromText(
text WKT, integer srid)
;
geometry ST_GeomCollFromText(
text WKT)
;
Makes a collection Geometry from the Well-Known-Text (WKT) representation with the given SRID. If SRID is not give, it defaults to 0.
OGC SPEC 3.2.6.2 - option SRID is from the conformance suite
Ritorna null se il WKT in input non è una GEOMETRYCOLLECTION
Se si è sicuri che tutti i WKT in input sono effettivamente GEOMETRYCOLLECTION, è sconsigliato usare questa funzione. È più lenta di ST_GeomFromText perché effettua anche una validazione della geometria. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
This method implements the SQL/MM specification.
ST_GeomFromEWKB — Ritorna un valore ST_Geometry a partire da Extended Well-Known Binary (EWKB).
geometry ST_GeomFromEWKB(
bytea EWKB)
;
Costruisce una ST_Geometry PostGIS a partire da OGC Extended Well-Known Binary (EWKT).
Il EWKB non è uno standard OGC ma uno specifico formato di Postgis che include il codice del sistema di riferimento spaziale (SRID). |
Miglioramento nella version 2.0.2: introdotto il supporto per superfici poliedriche e TIN.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
Rappresentazione binaria di LINESTRING(-71.160281 42.258729,-71.160837 42.259113,-71.161144 42.25932) in NAD 83 long lat (4269).
NOTE: Even though byte arrays are delimited with \ and may have ', we need to escape both out with \ and '' if standard_conforming_strings is off. So it does not look exactly like its AsEWKB representation. |
SELECT ST_GeomFromEWKB(E'\\001\\002\\000\\000 \\255\\020\\000\\000\\003\\000\\000\\000\\344J= \\013B\\312Q\\300n\\303(\\010\\036!E@''\\277E''K \\312Q\\300\\366{b\\235*!E@\\225|\\354.P\\312Q \\300p\\231\\323e1!E@');
In PostgreSQL 9.1+ il parametro standard_conforming_strings is settato su on per default, mentre nelle versioni precedenti era settato su off. Questa impostazione può essere cambiata per ogni singola query, a livello del database e a livello dell'intero server. Nell'esempio seguente si può vedere come costruire la query con standard_conforming_strings = on. .... |
set standard_conforming_strings = on; SELECT ST_GeomFromEWKB('\001\002\000\000 \255\020\000\000\003\000\000\000\344J=\012\013B \312Q\300n\303(\010\036!E@''\277E''K\012\312Q\300\366{b\235*!E@\225|\354.P\312Q\012\300p\231\323e1')
ST_GeomFromEWKT — Ritorna un valore ST_Geometry a partire da una rappresentazione Extended Well-Known Text (EWKT).
geometry ST_GeomFromEWKT(
text EWKT)
;
Costruisce un oggetto PostGIS ST_Geometry a partire da una rappresentazione OGC Extended Well-Known Text (EWKT).
Il formato EWKT non è uno standard OGC ma un formato specifico di PostGIS che include il codice del sistema di riferimento spaziale (SRID). |
Miglioramento nella version 2.0.2: introdotto il supporto per superfici poliedriche e TIN.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT ST_GeomFromEWKT('SRID=4269;LINESTRING(-71.160281 42.258729,-71.160837 42.259113,-71.161144 42.25932)'); SELECT ST_GeomFromEWKT('SRID=4269;MULTILINESTRING((-71.160281 42.258729,-71.160837 42.259113,-71.161144 42.25932))'); SELECT ST_GeomFromEWKT('SRID=4269;POINT(-71.064544 42.28787)'); SELECT ST_GeomFromEWKT('SRID=4269;POLYGON((-71.1776585052917 42.3902909739571,-71.1776820268866 42.3903701743239, -71.1776063012595 42.3903825660754,-71.1775826583081 42.3903033653531,-71.1776585052917 42.3902909739571))'); SELECT ST_GeomFromEWKT('SRID=4269;MULTIPOLYGON(((-71.1031880899493 42.3152774590236, -71.1031627617667 42.3152960829043,-71.102923838298 42.3149156848307, -71.1023097974109 42.3151969047397,-71.1019285062273 42.3147384934248, -71.102505233663 42.3144722937587,-71.10277487471 42.3141658254797, -71.103113945163 42.3142739188902,-71.10324876416 42.31402489987, -71.1033002961013 42.3140393340215,-71.1033488797549 42.3139495090772, -71.103396240451 42.3138632439557,-71.1041521907712 42.3141153348029, -71.1041411411543 42.3141545014533,-71.1041287795912 42.3142114839058, -71.1041188134329 42.3142693656241,-71.1041112482575 42.3143272556118, -71.1041072845732 42.3143851580048,-71.1041057218871 42.3144430686681, -71.1041065602059 42.3145009876017,-71.1041097995362 42.3145589148055, -71.1041166403905 42.3146168544148,-71.1041258822717 42.3146748022936, -71.1041375307579 42.3147318674446,-71.1041492906949 42.3147711126569, -71.1041598612795 42.314808571739,-71.1042515013869 42.3151287620809, -71.1041173835118 42.3150739481917,-71.1040809891419 42.3151344119048, -71.1040438678912 42.3151191367447,-71.1040194562988 42.3151832057859, -71.1038734225584 42.3151140942995,-71.1038446938243 42.3151006300338, -71.1038315271889 42.315094347535,-71.1037393329282 42.315054824985, -71.1035447555574 42.3152608696313,-71.1033436658644 42.3151648370544, -71.1032580383161 42.3152269126061,-71.103223066939 42.3152517403219, -71.1031880899493 42.3152774590236)), ((-71.1043632495873 42.315113108546,-71.1043583974082 42.3151211109857, -71.1043443253471 42.3150676015829,-71.1043850704575 42.3150793250568,-71.1043632495873 42.315113108546)))');
-- Linea curva 3d SELECT ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)');
--Esempio di superficie poliedrica SELECT ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )');
ST_GeometryFromText — Restituisce un valore ST_Geometry a partire da Well-Known-Text (WKT). È un alias per ST_GeomFromText
geometry ST_GeometryFromText(
text WKT)
;
geometry ST_GeometryFromText(
text WKT, integer srid)
;
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.40
ST_GeomFromGeoHash — Return a geometry from a GeoHash string.
geometry ST_GeomFromGeoHash(
text geohash, integer precision=full_precision_of_geohash)
;
Return a geometry from a GeoHash string. The geometry will be a polygon representing the GeoHash bounds.
If no precision
is specified ST_GeomFromGeoHash returns a polygon based on full precision of the input GeoHash string.
If precision
is specified ST_GeomFromGeoHash will use that many characters from the GeoHash to create the polygon.
Availability: 2.1.0
SELECT ST_AsText(ST_GeomFromGeoHash('9qqj7nmxncgyy4d0dbxqz0')); st_astext -------------------------------------------------------------------------------------------------------------------------- POLYGON((-115.172816 36.114646,-115.172816 36.114646,-115.172816 36.114646,-115.172816 36.114646,-115.172816 36.114646)) SELECT ST_AsText(ST_GeomFromGeoHash('9qqj7nmxncgyy4d0dbxqz0', 4)); st_astext ------------------------------------------------------------------------------------------------------------------------------ POLYGON((-115.3125 36.03515625,-115.3125 36.2109375,-114.9609375 36.2109375,-114.9609375 36.03515625,-115.3125 36.03515625)) SELECT ST_AsText(ST_GeomFromGeoHash('9qqj7nmxncgyy4d0dbxqz0', 10)); st_astext ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- POLYGON((-115.17282128334 36.1146408319473,-115.17282128334 36.1146461963654,-115.172810554504 36.1146461963654,-115.172810554504 36.1146408319473,-115.17282128334 36.1146408319473))
ST_GeomFromGML — Accetta una geometria in formato GML come input e restituisce un oggetto PostGIS geometry
geometry ST_GeomFromGML(
text geomgml)
;
geometry ST_GeomFromGML(
text geomgml, integer srid)
;
Costruisce un oggetto PostGIS ST_Geometry a partire da una rappresentazione OGC GML
ST_GeomFromGML supporta solo singole geometrie in formato GML. Ritorna un errore se si cerca di usare l'intero documento GML.
Versioni di OGC GML supportate:
GML 3.2.1 Namespace
GML 3.1.1 Simple Features profile SF-2 (con retrocompatibilità GML 3.1.0 e 3.0.0)
GML 2.1.2
OGC GML standards, cf: http://www.opengeospatial.org/standards/gml:
Disponibilità: 1.5, richiede libxml2 1.6+
Miglioramento nella version 2.0.2: introdotto il supporto per superfici poliedriche e TIN.
Miglioramento nella versione: 2.0.0 introdotto opzionale parametro SRID.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
In formato GML permette di combinare diverse dimensioni (per esempio la stessa multigeometria può contenere elementi in 2D e 3D contemporaneamente). Siccome le geometrie PostGIS non supportano la combinazione di diverse dimensioni, la funzione ST_GeomFromGML converte l'intera geometria in 2D se almeno un elemento è in 2D.
Il formato GML supporta la combinazione di diversi SRID all'interno della stessa multigeometria. Siccome le geometrie PostGIS non supportano questa combinazione, la funzione ST_GeomFromGML riproietta tutte le sottogeometrie nello SRID del nodo root. Se il nodo root non contiene l'attributo srsName, la funzione restituisce un errore.
La funzione ST_GeomFromGML non richiede dell'indicazione esplicita di un namespace GML. Nell'uso comune una indicazione specifica. non è necessaria. Il namespace deve però essere indicato esplicitamente se si vuole usare la feature XLink all'interno di GML.
ST_GeomFromGML non supporta geometrie curve SQL/MM. |
SELECT ST_GeomFromGML(' <gml:LineString srsName="EPSG:4269"> <gml:coordinates> -71.16028,42.258729 -71.160837,42.259112 -71.161143,42.25932 </gml:coordinates> </gml:LineString >');
SELECT ST_GeomFromGML(' <gml:LineString xmlns:gml="http://www.opengis.net/gml" xmlns:xlink="http://www.w3.org/1999/xlink" srsName="urn:ogc:def:crs:EPSG::4269"> <gml:pointProperty> <gml:Point gml:id="p1" ><gml:pos >42.258729 -71.16028</gml:pos ></gml:Point> </gml:pointProperty> <gml:pos >42.259112 -71.160837</gml:pos> <gml:pointProperty> <gml:Point xlink:type="simple" xlink:href="#p1"/> </gml:pointProperty> </gml:LineString >'););
SELECT ST_AsEWKT(ST_GeomFromGML(' <gml:PolyhedralSurface> <gml:polygonPatches> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing ><gml:posList srsDimension="3" >0 0 0 0 0 1 0 1 1 0 1 0 0 0 0</gml:posList ></gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing ><gml:posList srsDimension="3" >0 0 0 0 1 0 1 1 0 1 0 0 0 0 0</gml:posList ></gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing ><gml:posList srsDimension="3" >0 0 0 1 0 0 1 0 1 0 0 1 0 0 0</gml:posList ></gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing ><gml:posList srsDimension="3" >1 1 0 1 1 1 1 0 1 1 0 0 1 1 0</gml:posList ></gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing ><gml:posList srsDimension="3" >0 1 0 0 1 1 1 1 1 1 1 0 0 1 0</gml:posList ></gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing ><gml:posList srsDimension="3" >0 0 1 1 0 1 1 1 1 0 1 1 0 0 1</gml:posList ></gml:LinearRing> </gml:exterior> </gml:PolygonPatch> </gml:polygonPatches> </gml:PolyhedralSurface >')); -- risultato -- POLYHEDRALSURFACE(((0 0 0,0 0 1,0 1 1,0 1 0,0 0 0)), ((0 0 0,0 1 0,1 1 0,1 0 0,0 0 0)), ((0 0 0,1 0 0,1 0 1,0 0 1,0 0 0)), ((1 1 0,1 1 1,1 0 1,1 0 0,1 1 0)), ((0 1 0,0 1 1,1 1 1,1 1 0,0 1 0)), ((0 0 1,1 0 1,1 1 1,0 1 1,0 0 1)))
ST_GeomFromGeoJSON — Accetta come input la rappresentazione geojson di una geometria e restituisce una geometria PostGIS
geometry ST_GeomFromGeoJSON(
text geomjson)
;
Costruisce una geometria PostGIS a partire a una rappresentazione GeoJson
ST_GeomFromGeoJSON funzione solo con frammenti di geometrie GeoJson. Ritorna un errore se si cerca di usare un completo documento json.
Disponibilità: 2.0.0. Richiede - JSON-C >= 0.9
Se supporto JSON-C non è attivato, viene restituito un errore invece dell'output. Per attivare il supporto JSON-C, usare configure --with-jsondir=/path/to/json-c. Vedi Section 2.4.1, “Configurazione” per dettagli. |
This function supports 3d and will not drop the z-index.
SELECT ST_AsText(ST_GeomFromGeoJSON('{"type":"Point","coordinates":[-48.23456,20.12345]}')) As wkt; wkt ------ POINT(-48.23456 20.12345)
-- una linestring 3D SELECT ST_AsText(ST_GeomFromGeoJSON('{"type":"LineString","coordinates":[[1,2,3],[4,5,6],[7,8,9]]}')) As wkt; wkt ------------------- LINESTRING(1 2,4 5,7 8)
ST_GeomFromKML — Accetta come input una geometria in formato KML e restituisce una geometria PostGIS.
geometry ST_GeomFromKML(
text geomkml)
;
Costruisce un oggetto PostSIG ST_Geometry a partire da una geometria in formato OGC KML.
ST_GeomFromKML accetta solo frammenti di geometrie KML. Restituisce un errore se l'input consiste in un intero documento KML.
Versioni di OGC KML supportate:
KML 2.2.0 Namespace
OGC KML standards, cf: http://www.opengeospatial.org/standards/kml:
Disponibilità: PostGIS 1.5, libxml2 2.6+
This function supports 3d and will not drop the z-index.
La funzione ST_GeomFromKML non supporta geometrie curve SQL/MM. |
ST_GMLToSQL — Restituisce un valore ST_Geometry a partire da una rappresentazione GML. Questo è solo un alias per la funzione ST_GeomFromGML.
geometry ST_GMLToSQL(
text geomgml)
;
geometry ST_GMLToSQL(
text geomgml, integer srid)
;
This method implements the SQL/MM specification. SQL-MM 3: 5.1.50 (tranne che il supporto per le curve).
Disponibilità: 1.5, richiede libxml2 1.6+
Miglioramento nella version 2.0.2: introdotto il supporto per superfici poliedriche e TIN.
Miglioramento nella versione: 2.0.0 introdotto opzionale parametro SRID.
ST_GeomFromText — Restituisce un valore ST_Geometry a partire da una rappresentazione Well-Known-Text (WKT)
geometry ST_GeomFromText(
text WKT)
;
geometry ST_GeomFromText(
text WKT, integer srid)
;
Costruisce un oggetto PostGIS ST_Geometry a partire da geometria in formato OGC Well-Known-Text
There are 2 variants of ST_GeomFromText function, the first takes no SRID and returns a geometry with no defined spatial reference system. The second takes a spatial reference id as the second argument and returns an ST_Geometry that includes this srid as part of its meta-data. The srid must be defined in the spatial_ref_sys table. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2 - option SRID is from the conformance suite.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.40
This method supports Circular Strings and Curves
Changed: 2.0.0 In prior versions of PostGIS ST_GeomFromText('GEOMETRYCOLLECTION(EMPTY)') was allowed. This is now illegal in PostGIS 2.0.0 to better conform with SQL/MM standards. This should now be written as ST_GeomFromText('GEOMETRYCOLLECTION EMPTY') |
SELECT ST_GeomFromText('LINESTRING(-71.160281 42.258729,-71.160837 42.259113,-71.161144 42.25932)'); SELECT ST_GeomFromText('LINESTRING(-71.160281 42.258729,-71.160837 42.259113,-71.161144 42.25932)',4269); SELECT ST_GeomFromText('MULTILINESTRING((-71.160281 42.258729,-71.160837 42.259113,-71.161144 42.25932))'); SELECT ST_GeomFromText('POINT(-71.064544 42.28787)'); SELECT ST_GeomFromText('POLYGON((-71.1776585052917 42.3902909739571,-71.1776820268866 42.3903701743239, -71.1776063012595 42.3903825660754,-71.1775826583081 42.3903033653531,-71.1776585052917 42.3902909739571))'); SELECT ST_GeomFromText('MULTIPOLYGON(((-71.1031880899493 42.3152774590236, -71.1031627617667 42.3152960829043,-71.102923838298 42.3149156848307, -71.1023097974109 42.3151969047397,-71.1019285062273 42.3147384934248, -71.102505233663 42.3144722937587,-71.10277487471 42.3141658254797, -71.103113945163 42.3142739188902,-71.10324876416 42.31402489987, -71.1033002961013 42.3140393340215,-71.1033488797549 42.3139495090772, -71.103396240451 42.3138632439557,-71.1041521907712 42.3141153348029, -71.1041411411543 42.3141545014533,-71.1041287795912 42.3142114839058, -71.1041188134329 42.3142693656241,-71.1041112482575 42.3143272556118, -71.1041072845732 42.3143851580048,-71.1041057218871 42.3144430686681, -71.1041065602059 42.3145009876017,-71.1041097995362 42.3145589148055, -71.1041166403905 42.3146168544148,-71.1041258822717 42.3146748022936, -71.1041375307579 42.3147318674446,-71.1041492906949 42.3147711126569, -71.1041598612795 42.314808571739,-71.1042515013869 42.3151287620809, -71.1041173835118 42.3150739481917,-71.1040809891419 42.3151344119048, -71.1040438678912 42.3151191367447,-71.1040194562988 42.3151832057859, -71.1038734225584 42.3151140942995,-71.1038446938243 42.3151006300338, -71.1038315271889 42.315094347535,-71.1037393329282 42.315054824985, -71.1035447555574 42.3152608696313,-71.1033436658644 42.3151648370544, -71.1032580383161 42.3152269126061,-71.103223066939 42.3152517403219, -71.1031880899493 42.3152774590236)), ((-71.1043632495873 42.315113108546,-71.1043583974082 42.3151211109857, -71.1043443253471 42.3150676015829,-71.1043850704575 42.3150793250568,-71.1043632495873 42.315113108546)))',4326); SELECT ST_GeomFromText('CIRCULARSTRING(220268 150415,220227 150505,220227 150406)');
ST_GeomFromWKB — Creates a geometry instance from a Well-Known Binary geometry representation (WKB) and optional SRID.
geometry ST_GeomFromWKB(
bytea geom)
;
geometry ST_GeomFromWKB(
bytea geom, integer srid)
;
The ST_GeomFromWKB
function, takes a well-known binary representation of a geometry and a Spatial Reference System ID (SRID
) and creates an instance of the appropriate geometry type. This function plays the role of the Geometry Factory in SQL. This is an alternate name for ST_WKBToSQL.
If SRID is not specified, it defaults to 0 (Unknown).
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.7.2 - the optional SRID is from the conformance suite
This method implements the SQL/MM specification. SQL-MM 3: 5.1.41
This method supports Circular Strings and Curves
--Although bytea rep contains single \, these need to be escaped when inserting into a table -- unless standard_conforming_strings is set to on. SELECT ST_AsEWKT( ST_GeomFromWKB(E'\\001\\002\\000\\000\\000\\002\\000\\000\\000\\037\\205\\353Q\\270~\\\\\\300\\323Mb\\020X\\231C@\\020X9\\264\\310~\\\\\\300)\\\\\\217\\302\\365\\230C@',4326) ); st_asewkt ------------------------------------------------------ SRID=4326;LINESTRING(-113.98 39.198,-113.981 39.195) (1 row) SELECT ST_AsText( ST_GeomFromWKB( ST_AsEWKB('POINT(2 5)'::geometry) ) ); st_astext ------------ POINT(2 5) (1 row)
ST_LineFromEncodedPolyline — Creates a LineString from an Encoded Polyline.
geometry ST_LineFromEncodedPolyline(
text polyline, integer precision=5)
;
Creates a LineString from an Encoded Polyline string.
See http://developers.google.com/maps/documentation/utilities/polylinealgorithm
Disponibilità: PostGIS 1.5, libxml2 2.6+
ST_LineFromMultiPoint — Crea una LineString da una geometria MultiPoint.
geometry ST_LineFromMultiPoint(
geometry aMultiPoint)
;
Crea una LineString da una geometria MultiPoint.
This function supports 3d and will not drop the z-index.
ST_LineFromText — Makes a Geometry from WKT representation with the given SRID. If SRID is not given, it defaults to 0.
geometry ST_LineFromText(
text WKT)
;
geometry ST_LineFromText(
text WKT, integer srid)
;
Makes a Geometry from WKT with the given SRID. If SRID is not give, it defaults to 0. If WKT passed in is not a LINESTRING, then null is returned.
OGC SPEC 3.2.6.2 - option SRID is from the conformance suite. |
If you know all your geometries are LINESTRINGS, its more efficient to just use ST_GeomFromText. This just calls ST_GeomFromText and adds additional validation that it returns a linestring. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
This method implements the SQL/MM specification. SQL-MM 3: 7.2.8
ST_LineFromWKB — Makes a LINESTRING
from WKB with the given SRID
geometry ST_LineFromWKB(
bytea WKB)
;
geometry ST_LineFromWKB(
bytea WKB, integer srid)
;
The ST_LineFromWKB
function, takes a well-known binary representation of geometry and a Spatial Reference System ID (SRID
) and creates an instance of the appropriate geometry type - in this case, a LINESTRING
geometry. This function plays the role of the Geometry Factory in SQL.
If an SRID is not specified, it defaults to 0. NULL
is returned if the input bytea
does not represent a LINESTRING
.
OGC SPEC 3.2.6.2 - option SRID is from the conformance suite. |
If you know all your geometries are |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
This method implements the SQL/MM specification. SQL-MM 3: 7.2.9
ST_LinestringFromWKB — Makes a geometry from WKB with the given SRID.
geometry ST_LinestringFromWKB(
bytea WKB)
;
geometry ST_LinestringFromWKB(
bytea WKB, integer srid)
;
The ST_LinestringFromWKB
function, takes a well-known binary representation of geometry and a Spatial Reference System ID (SRID
) and creates an instance of the appropriate geometry type - in this case, a LINESTRING
geometry. This function plays the role of the Geometry Factory in SQL.
If an SRID is not specified, it defaults to 0. NULL
is returned if the input bytea
does not represent a LINESTRING
geometry. This an alias for ST_LineFromWKB.
OGC SPEC 3.2.6.2 - optional SRID is from the conformance suite. |
If you know all your geometries are |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
This method implements the SQL/MM specification. SQL-MM 3: 7.2.9
ST_MakeBox2D — Crea una BOX2D definita dalle geometrie dei punti dati.
box2d ST_MakeBox2D(
geometry pointLowLeft, geometry pointUpRight)
;
Creates a BOX2D defined by the given point geometries. This is useful for doing range queries
--Return all features that fall reside or partly reside in a US national atlas coordinate bounding box --It is assumed here that the geometries are stored with SRID = 2163 (US National atlas equal area) SELECT feature_id, feature_name, the_geom FROM features WHERE the_geom && ST_SetSRID(ST_MakeBox2D(ST_Point(-989502.1875, 528439.5625), ST_Point(-987121.375 ,529933.1875)),2163)
ST_3DMakeBox — Crea una BOX3D definita dalle geometrie dei punti 3d dati.
box3d ST_3DMakeBox(
geometry point3DLowLeftBottom, geometry point3DUpRightTop)
;
Creates a BOX3D defined by the given 2 3D point geometries.
This function supports 3d and will not drop the z-index.
Changed: 2.0.0 In prior versions this used to be called ST_MakeBox3D
ST_MakeLine — Creates a Linestring from point or line geometries.
geometry ST_MakeLine(
geometry set geoms)
;
geometry ST_MakeLine(
geometry geom1, geometry geom2)
;
geometry ST_MakeLine(
geometry[] geoms_array)
;
ST_MakeLine comes in 3 forms: a spatial aggregate that takes rows of point-or-line geometries and returns a line string, a function that takes an array of point-or-lines, and a regular function that takes two point-or-line geometries. You might want to use a subselect to order points before feeding them to the aggregate version of this function.
When adding line components a common node is removed from the output.
This function supports 3d and will not drop the z-index.
Availability: 1.4.0 - ST_MakeLine(geomarray) was introduced. ST_MakeLine aggregate functions was enhanced to handle more points faster.
Availability: 2.0.0 - Support for linestring input elements was introduced
This example takes a sequence of GPS points and creates one record for each gps travel where the geometry field is a line string composed of the gps points in the order of the travel.
-- For pre-PostgreSQL 9.0 - this usually works, -- but the planner may on occasion choose not to respect the order of the subquery SELECT gps.gps_track, ST_MakeLine(gps.the_geom) As newgeom FROM (SELECT gps_track,gps_time, the_geom FROM gps_points ORDER BY gps_track, gps_time) As gps GROUP BY gps.gps_track;
-- If you are using PostgreSQL 9.0+ -- (you can use the new ORDER BY support for aggregates) -- this is a guaranteed way to get a correctly ordered linestring -- Your order by part can order by more than one column if needed SELECT gps.gps_track, ST_MakeLine(gps.the_geom ORDER BY gps_time) As newgeom FROM gps_points As gps GROUP BY gps.gps_track;
First example is a simple one off line string composed of 2 points. The second formulates line strings from 2 points a user draws. The third is a one-off that joins 2 3d points to create a line in 3d space.
SELECT ST_AsText(ST_MakeLine(ST_MakePoint(1,2), ST_MakePoint(3,4))); st_astext --------------------- LINESTRING(1 2,3 4) SELECT userpoints.id, ST_MakeLine(startpoint, endpoint) As drawn_line FROM userpoints ; SELECT ST_AsEWKT(ST_MakeLine(ST_MakePoint(1,2,3), ST_MakePoint(3,4,5))); st_asewkt ------------------------- LINESTRING(1 2 3,3 4 5)
SELECT ST_MakeLine(ARRAY(SELECT ST_Centroid(the_geom) FROM visit_locations ORDER BY visit_time)); --Making a 3d line with 3 3-d points SELECT ST_AsEWKT(ST_MakeLine(ARRAY[ST_MakePoint(1,2,3), ST_MakePoint(3,4,5), ST_MakePoint(6,6,6)])); st_asewkt ------------------------- LINESTRING(1 2 3,3 4 5,6 6 6)
ST_MakeEnvelope — Creates a rectangular Polygon formed from the given minimums and maximums. Input values must be in SRS specified by the SRID.
geometry ST_MakeEnvelope(
double precision xmin, double precision ymin, double precision xmax, double precision ymax, integer srid=unknown)
;
Creates a rectangular Polygon formed from the minima and maxima. by the given shell. Input values must be in SRS specified by the SRID. If no SRID is specified the unknown spatial reference system is assumed
Availability: 1.5
Enhanced: 2.0: Ability to specify an envelope without specifying an SRID was introduced.
ST_MakePolygon — Creates a Polygon formed by the given shell. Input geometries must be closed LINESTRINGS.
geometry ST_MakePolygon(
geometry linestring)
;
geometry ST_MakePolygon(
geometry outerlinestring, geometry[] interiorlinestrings)
;
Creates a Polygon formed by the given shell. Input geometries must be closed LINESTRINGS. Comes in 2 variants.
Variant 1: takes one closed linestring.
Variant 2: Creates a Polygon formed by the given shell and array of holes. You can construct a geometry array using ST_Accum or the PostgreSQL ARRAY[] and ARRAY() constructs. Input geometries must be closed LINESTRINGS.
This function will not accept a MULTILINESTRING. Use ST_LineMerge or ST_Dump to generate line strings. |
This function supports 3d and will not drop the z-index.
--linea 2d SELECT ST_MakePolygon(ST_GeomFromText('LINESTRING(75.15 29.53,77 29,77.6 29.5, 75.15 29.53)')); --se la linestring non è chiusa --potete aggiungere il punto iniziale per chiuderla SELECT ST_MakePolygon(ST_AddPoint(foo.open_line, ST_StartPoint(foo.open_line))) FROM ( SELECT ST_GeomFromText('LINESTRING(75.15 29.53,77 29,77.6 29.5)') As open_line) As foo; --linea 3d chiusa SELECT ST_MakePolygon(ST_GeomFromText('LINESTRING(75.15 29.53 1,77 29 1,77.6 29.5 1, 75.15 29.53 1)')); st_asewkt ----------- POLYGON((75.15 29.53 1,77 29 1,77.6 29.5 1,75.15 29.53 1)) --linea misurata -- SELECT ST_MakePolygon(ST_GeomFromText('LINESTRINGM(75.15 29.53 1,77 29 1,77.6 29.5 2, 75.15 29.53 2)')); st_asewkt ---------- POLYGONM((75.15 29.53 1,77 29 1,77.6 29.5 2,75.15 29.53 2))
Build a donut with an ant hole
SELECT ST_MakePolygon( ST_ExteriorRing(ST_Buffer(foo.line,10)), ARRAY[ST_Translate(foo.line,1,1), ST_ExteriorRing(ST_Buffer(ST_MakePoint(20,20),1)) ] ) FROM (SELECT ST_ExteriorRing(ST_Buffer(ST_MakePoint(10,10),10,10)) As line ) As foo;
Build province boundaries with holes representing lakes in the province from a set of province polygons/multipolygons and water line strings this is an example of using PostGIS ST_Accum
The use of CASE because feeding a null array into ST_MakePolygon results in NULL |
the use of left join to guarantee we get all provinces back even if they have no lakes |
SELECT p.gid, p.province_name, CASE WHEN ST_Accum(w.the_geom) IS NULL THEN p.the_geom ELSE ST_MakePolygon(ST_LineMerge(ST_Boundary(p.the_geom)), ST_Accum(w.the_geom)) END FROM provinces p LEFT JOIN waterlines w ON (ST_Within(w.the_geom, p.the_geom) AND ST_IsClosed(w.the_geom)) GROUP BY p.gid, p.province_name, p.the_geom; -- Stesso esempio di cui sopra, ma con una sottoquery correlata -- e la funzione PostgreSQL ARRAY() che converte il set di record in un vettore SELECT p.gid, p.province_name, CASE WHEN EXISTS(SELECT w.the_geom FROM waterlines w WHERE ST_Within(w.the_geom, p.the_geom) AND ST_IsClosed(w.the_geom)) THEN ST_MakePolygon(ST_LineMerge(ST_Boundary(p.the_geom)), ARRAY(SELECT w.the_geom FROM waterlines w WHERE ST_Within(w.the_geom, p.the_geom) AND ST_IsClosed(w.the_geom))) ELSE p.the_geom END As the_geom FROM provinces p;
ST_Boundary, ST_Accum, ST_AddPoint, ST_GeometryType, ST_IsClosed, ST_LineMerge, ST_BuildArea
ST_MakePoint — Creates a 2D,3DZ or 4D point geometry.
geometry ST_MakePoint(
double precision x, double precision y)
;
geometry ST_MakePoint(
double precision x, double precision y, double precision z)
;
geometry ST_MakePoint(
double precision x, double precision y, double precision z, double precision m)
;
Creates a 2D,3DZ or 4D point geometry (geometry with measure). ST_MakePoint
while not being OGC compliant is generally faster and more precise than ST_GeomFromText and ST_PointFromText. It is also easier to use if you have raw coordinates rather than WKT.
Note x is longitude and y is latitude |
Utilizzare ST_MakePointM se dovete creare un punto con x,y,m. |
This function supports 3d and will not drop the z-index.
--Return point with unknown SRID SELECT ST_MakePoint(-71.1043443253471, 42.3150676015829); --Return point marked as WGS 84 long lat SELECT ST_SetSRID(ST_MakePoint(-71.1043443253471, 42.3150676015829),4326); --Return a 3D point (e.g. has altitude) SELECT ST_MakePoint(1, 2,1.5); --Get z of point SELECT ST_Z(ST_MakePoint(1, 2,1.5)); result ------- 1.5
ST_MakePointM — Creates a point geometry with an x y and m coordinate.
geometry ST_MakePointM(
float x, float y, float m)
;
Creates a point with x, y and measure coordinates.
Note x is longitude and y is latitude. |
We use ST_AsEWKT in these examples to show the text representation instead of ST_AsText because ST_AsText does not support returning M.
--Return EWKT representation of point with unknown SRID SELECT ST_AsEWKT(ST_MakePointM(-71.1043443253471, 42.3150676015829, 10)); --result st_asewkt ----------------------------------------------- POINTM(-71.1043443253471 42.3150676015829 10) --Return EWKT representation of point with measure marked as WGS 84 long lat SELECT ST_AsEWKT(ST_SetSRID(ST_MakePointM(-71.1043443253471, 42.3150676015829,10),4326)); st_asewkt --------------------------------------------------------- SRID=4326;POINTM(-71.1043443253471 42.3150676015829 10) --Return a 3d point (e.g. has altitude) SELECT ST_MakePoint(1, 2,1.5); --Get m of point SELECT ST_M(ST_MakePointM(-71.1043443253471, 42.3150676015829,10)); result ------- 10
ST_MLineFromText — Return a specified ST_MultiLineString value from WKT representation.
geometry ST_MLineFromText(
text WKT, integer srid)
;
geometry ST_MLineFromText(
text WKT)
;
Makes a Geometry from Well-Known-Text (WKT) with the given SRID. If SRID is not give, it defaults to 0.
OGC SPEC 3.2.6.2 - option SRID is from the conformance suite
Returns null if the WKT is not a MULTILINESTRING
If you are absolutely sure all your WKT geometries are points, don't use this function. It is slower than ST_GeomFromText since it adds an additional validation step. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
This method implements the SQL/MM specification.SQL-MM 3: 9.4.4
ST_MPointFromText — Makes a Geometry from WKT with the given SRID. If SRID is not give, it defaults to 0.
geometry ST_MPointFromText(
text WKT, integer srid)
;
geometry ST_MPointFromText(
text WKT)
;
Makes a Geometry from WKT with the given SRID. If SRID is not give, it defaults to 0.
OGC SPEC 3.2.6.2 - option SRID is from the conformance suite
Returns null if the WKT is not a MULTIPOINT
If you are absolutely sure all your WKT geometries are points, don't use this function. It is slower than ST_GeomFromText since it adds an additional validation step. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. 3.2.6.2
This method implements the SQL/MM specification. SQL-MM 3: 9.2.4
ST_MPolyFromText — Makes a MultiPolygon Geometry from WKT with the given SRID. If SRID is not give, it defaults to 0.
geometry ST_MPolyFromText(
text WKT, integer srid)
;
geometry ST_MPolyFromText(
text WKT)
;
Makes a MultiPolygon from WKT with the given SRID. If SRID is not give, it defaults to 0.
OGC SPEC 3.2.6.2 - option SRID is from the conformance suite
Throws an error if the WKT is not a MULTIPOLYGON
If you are absolutely sure all your WKT geometries are multipolygons, don't use this function. It is slower than ST_GeomFromText since it adds an additional validation step. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
This method implements the SQL/MM specification. SQL-MM 3: 9.6.4
SELECT ST_MPolyFromText('MULTIPOLYGON(((0 0 1,20 0 1,20 20 1,0 20 1,0 0 1),(5 5 3,5 7 3,7 7 3,7 5 3,5 5 3)))'); SELECt ST_MPolyFromText('MULTIPOLYGON(((-70.916 42.1002,-70.9468 42.0946,-70.9765 42.0872,-70.9754 42.0875,-70.9749 42.0879,-70.9752 42.0881,-70.9754 42.0891,-70.9758 42.0894,-70.9759 42.0897,-70.9759 42.0899,-70.9754 42.0902,-70.9756 42.0906,-70.9753 42.0907,-70.9753 42.0917,-70.9757 42.0924,-70.9755 42.0928,-70.9755 42.0942,-70.9751 42.0948,-70.9755 42.0953,-70.9751 42.0958,-70.9751 42.0962,-70.9759 42.0983,-70.9767 42.0987,-70.9768 42.0991,-70.9771 42.0997,-70.9771 42.1003,-70.9768 42.1005,-70.977 42.1011,-70.9766 42.1019,-70.9768 42.1026,-70.9769 42.1033,-70.9775 42.1042,-70.9773 42.1043,-70.9776 42.1043,-70.9778 42.1048,-70.9773 42.1058,-70.9774 42.1061,-70.9779 42.1065,-70.9782 42.1078,-70.9788 42.1085,-70.9798 42.1087,-70.9806 42.109,-70.9807 42.1093,-70.9806 42.1099,-70.9809 42.1109,-70.9808 42.1112,-70.9798 42.1116,-70.9792 42.1127,-70.979 42.1129,-70.9787 42.1134,-70.979 42.1139,-70.9791 42.1141,-70.9987 42.1116,-71.0022 42.1273, -70.9408 42.1513,-70.9315 42.1165,-70.916 42.1002)))',4326);
ST_Point — Returns an ST_Point with the given coordinate values. OGC alias for ST_MakePoint.
geometry ST_Point(
float x_lon, float y_lat)
;
Returns an ST_Point with the given coordinate values. MM compliant alias for ST_MakePoint that takes just an x and y.
This method implements the SQL/MM specification. SQL-MM 3: 6.1.2
SELECT CAST(ST_SetSRID(ST_Point(-71.1043443253471, 42.3150676015829),4326) As geography);
-- lo :: è l'abbreviazione PostgreSQL per il casting. SELECT ST_SetSRID(ST_Point(-71.1043443253471, 42.3150676015829),4326)::geography;
--If your point coordinates are in a different spatial reference from WGS-84 long lat, then you need to transform before casting -- This example we convert a point in Pennsylvania State Plane feet to WGS 84 and then geography SELECT ST_Transform(ST_SetSRID(ST_Point(3637510, 3014852),2273),4326)::geography;
ST_PointFromGeoHash — Return a point from a GeoHash string.
point ST_PointFromGeoHash(
text geohash, integer precision=full_precision_of_geohash)
;
Return a point from a GeoHash string. The point represents the center point of the GeoHash.
If no precision
is specified ST_PointFromGeoHash returns a point based on full precision of the input GeoHash string.
If precision
is specified ST_PointFromGeoHash will use that many characters from the GeoHash to create the point.
Availability: 2.1.0
SELECT ST_AsText(ST_PointFromGeoHash('9qqj7nmxncgyy4d0dbxqz0')); st_astext ------------------------------ POINT(-115.172816 36.114646) SELECT ST_AsText(ST_PointFromGeoHash('9qqj7nmxncgyy4d0dbxqz0', 4)); st_astext ----------------------------------- POINT(-115.13671875 36.123046875) SELECT ST_AsText(ST_PointFromGeoHash('9qqj7nmxncgyy4d0dbxqz0', 10)); st_astext ------------------------------------------- POINT(-115.172815918922 36.1146435141563)
ST_PointFromText — Makes a point Geometry from WKT with the given SRID. If SRID is not given, it defaults to unknown.
geometry ST_PointFromText(
text WKT)
;
geometry ST_PointFromText(
text WKT, integer srid)
;
Constructs a PostGIS ST_Geometry point object from the OGC Well-Known text representation. If SRID is not give, it defaults to unknown (currently 0). If geometry is not a WKT point representation, returns null. If completely invalid WKT, then throws an error.
There are 2 variants of ST_PointFromText function, the first takes no SRID and returns a geometry with no defined spatial reference system. The second takes a spatial reference id as the second argument and returns an ST_Geometry that includes this srid as part of its meta-data. The srid must be defined in the spatial_ref_sys table. |
If you are absolutely sure all your WKT geometries are points, don't use this function. It is slower than ST_GeomFromText since it adds an additional validation step. If you are building points from long lat coordinates and care more about performance and accuracy than OGC compliance, use ST_MakePoint or OGC compliant alias ST_Point. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2 - option SRID is from the conformance suite.
This method implements the SQL/MM specification. SQL-MM 3: 6.1.8
ST_PointFromWKB — Makes a geometry from WKB with the given SRID
geometry ST_GeomFromWKB(
bytea geom)
;
geometry ST_GeomFromWKB(
bytea geom, integer srid)
;
The ST_PointFromWKB
function, takes a well-known binary representation of geometry and a Spatial Reference System ID (SRID
) and creates an instance of the appropriate geometry type - in this case, a POINT
geometry. This function plays the role of the Geometry Factory in SQL.
If an SRID is not specified, it defaults to 0. NULL
is returned if the input bytea
does not represent a POINT
geometry.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.7.2
This method implements the SQL/MM specification. SQL-MM 3: 6.1.9
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
ST_Polygon — Returns a polygon built from the specified linestring and SRID.
geometry ST_Polygon(
geometry aLineString, integer srid)
;
Returns a polygon built from the specified linestring and SRID.
ST_Polygon is similar to first version oST_MakePolygon except it also sets the spatial ref sys (SRID) of the polygon. Will not work with MULTILINESTRINGS so use LineMerge to merge multilines. Also does not create polygons with holes. Use ST_MakePolygon for that. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 8.3.2
This function supports 3d and will not drop the z-index.
--un poligono 2d SELECT ST_Polygon(ST_GeomFromText('LINESTRING(75.15 29.53,77 29,77.6 29.5, 75.15 29.53)'), 4326); --risultato-- POLYGON((75.15 29.53,77 29,77.6 29.5,75.15 29.53)) --un poligono 3d SELECT ST_AsEWKT(ST_Polygon(ST_GeomFromEWKT('LINESTRING(75.15 29.53 1,77 29 1,77.6 29.5 1, 75.15 29.53 1)'), 4326)); risultato ------ SRID=4326;POLYGON((75.15 29.53 1,77 29 1,77.6 29.5 1,75.15 29.53 1))
ST_PolygonFromText — Makes a Geometry from WKT with the given SRID. If SRID is not give, it defaults to 0.
geometry ST_PolygonFromText(
text WKT)
;
geometry ST_PolygonFromText(
text WKT, integer srid)
;
Makes a Geometry from WKT with the given SRID. If SRID is not give, it defaults to 0. Returns null if WKT is not a polygon.
OGC SPEC 3.2.6.2 - option SRID is from the conformance suite
If you are absolutely sure all your WKT geometries are polygons, don't use this function. It is slower than ST_GeomFromText since it adds an additional validation step. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.6.2
This method implements the SQL/MM specification. SQL-MM 3: 8.3.6
SELECT ST_PolygonFromText('POLYGON((-71.1776585052917 42.3902909739571,-71.1776820268866 42.3903701743239, -71.1776063012595 42.3903825660754,-71.1775826583081 42.3903033653531,-71.1776585052917 42.3902909739571))'); st_polygonfromtext ------------------ 010300000001000000050000006... SELECT ST_PolygonFromText('POINT(1 2)') IS NULL as point_is_notpoly; point_is_not_poly ---------- t
ST_WKBToSQL — Return a specified ST_Geometry value from Well-Known Binary representation (WKB). This is an alias name for ST_GeomFromWKB that takes no srid
geometry ST_WKBToSQL(
bytea WKB)
;
LINESTRING
or CIRCULARLINESTRING
geometry as a POINT
.POLYGON
. Restituisce NULL se la geometria non è un poligono. Non funziona con MULTIPOLYGONTRUE
se il punto iniziale e quello finale di LINESTRING
coincidono. Per le superfici poliedriche indica una superficie chiusa (volumetrica). TRUE
se l'argomento è una collezione (MULTI*
, GEOMETRYCOLLECTION
, ...) TRUE
if this LINESTRING
is both closed and simple.true
if the ST_Geometry
is well formed. LINESTRING
geometry as a POINT
.GeometryType — Restituisce il tipo di geometria come stringa. Per esempio: 'LINESTRING', 'POLYGON', 'MULTIPOINT', ecc.
text GeometryType(
geometry geomA)
;
Restituisce il tipo di geometria come stringa. Per esempio: 'LINESTRING', 'POLYGON', 'MULTIPOINT', ecc.
OGC SPEC s2.1.1.1 - Returns the name of the instantiable subtype of Geometry of which this Geometry instance is a member. The name of the instantiable subtype of Geometry is returned as a string.
Questa funzione indica anche se la geometria viene misurata, restituendo una stringa nella forma 'POINTM'. |
Migliorato: Nella 2.0.0 è stato introdotto il supporto per le superfici poliedriche, i triangoli e i TIN.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method supports Circular Strings and Curves
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT GeometryType(ST_GeomFromText('LINESTRING(77.29 29.07,77.42 29.26,77.27 29.31,77.29 29.07)')); geometrytype -------------- LINESTRING
SELECT ST_GeometryType(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')); --result POLYHEDRALSURFACE
SELECT GeometryType(geom) as result FROM (SELECT ST_GeomFromEWKT('TIN ((( 0 0 0, 0 0 1, 0 1 0, 0 0 0 )), (( 0 0 0, 0 1 0, 1 1 0, 0 0 0 )) )') AS geom ) AS g; result -------- TIN
ST_Boundary — Returns the closure of the combinatorial boundary of this Geometry.
geometry ST_Boundary(
geometry geomA)
;
Returns the closure of the combinatorial boundary of this Geometry. The combinatorial boundary is defined as described in section 3.12.3.2 of the OGC SPEC. Because the result of this function is a closure, and hence topologically closed, the resulting boundary can be represented using representational geometry primitives as discussed in the OGC SPEC, section 3.12.2.
Eseguito dal modulo GEOS
Prma della 2.0.0, questa funzione dava un'eccezione se usata con |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. OGC SPEC s2.1.1.1
This method implements the SQL/MM specification. SQL-MM 3: 5.1.14
This function supports 3d and will not drop the z-index.
Enhanced: 2.1.0 support for Triangle was introduced
SELECT ST_Boundary(geom) FROM (SELECT 'LINESTRING(100 150,50 60, 70 80, 160 170)'::geometry As geom) As f;
-- ST_AsText output MULTIPOINT(100 150,160 170)
|
SELECT ST_Boundary(geom) FROM (SELECT 'POLYGON (( 10 130, 50 190, 110 190, 140 150, 150 80, 100 10, 20 40, 10 130 ), ( 70 40, 100 50, 120 80, 80 110, 50 90, 70 40 ))'::geometry As geom) As f;
-- ST_AsText output MULTILINESTRING((10 130,50 190,110 190,140 150,150 80,100 10,20 40,10 130), (70 40,100 50,120 80,80 110,50 90,70 40))
|
SELECT ST_AsText(ST_Boundary(ST_GeomFromText('LINESTRING(1 1,0 0, -1 1)'))); st_astext ----------- MULTIPOINT(1 1,-1 1) SELECT ST_AsText(ST_Boundary(ST_GeomFromText('POLYGON((1 1,0 0, -1 1, 1 1))'))); st_astext ---------- LINESTRING(1 1,0 0,-1 1,1 1) --Usando un poligono 3d SELECT ST_AsEWKT(ST_Boundary(ST_GeomFromEWKT('POLYGON((1 1 1,0 0 1, -1 1 1, 1 1 1))'))); st_asewkt ----------------------------------- LINESTRING(1 1 1,0 0 1,-1 1 1,1 1 1) --Usando una multilinestring 3d SELECT ST_AsEWKT(ST_Boundary(ST_GeomFromEWKT('MULTILINESTRING((1 1 1,0 0 0.5, -1 1 1),(1 1 0.5,0 0 0.5, -1 1 0.5, 1 1 0.5) )'))); st_asewkt ---------- MULTIPOINT(-1 1 1,1 1 0.75)
ST_CoordDim — Restituisce la dimensione delle coordinate di ST_Geometry value.
integer ST_CoordDim(
geometry geomA)
;
Restituisce la dimensione delle coordinate di ST_Geometry value.
This is the MM compliant alias name for ST_NDims
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.3
This method supports Circular Strings and Curves
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_Dimension — La dimensione inerente a questo oggetto Geometry, che deve essere minore o uguale alla dimensione delle coordinate.
integer ST_Dimension(
geometry g)
;
La dimensione inerente a questo oggetto Geometry, che deve essere minore o uguale alla dimensione delle coordinate. La specifica OGC s2.1.1.1 restituisce 0 per POINT
, 1 per LINESTRING
, 2 per POLYGON
, e la maggiore fra le dimensioni dei componenti di una GEOMETRYCOLLECTION
. Se la geometria è sconosciuta (vuota), viene restituito il valore NULL.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.2
Miglioramento: nella versione 2.0.0 è stato introdotto il supporto per le supefici poliedriche e i TIN. Non viene più generata un'eccezione se viene data una geometria vuota.
Prima della 2.0.0, questa funzione dava un'eccezione se usata con una geometria vuota. |
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_EndPoint — Returns the last point of a LINESTRING
or CIRCULARLINESTRING
geometry as a POINT
.
boolean ST_EndPoint(
geometry g)
;
Restituisce l'ultimo punto di una geometria LINESTRING
come POINT
o NULL
se il parametro di input non è una LINESTRING
.
This method implements the SQL/MM specification. SQL-MM 3: 7.1.4
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
Modifica: La versione 2.0.0 non funziona più con geometrie singole di stringhe multilinea. Nelle versioni precedenti di PostGIS una stringa multilinea con una sola linea avrebbe funzionato tranquillamente con questa funzione, restituendo il punto di inizio. Nella versione 2.0.0 la funzione restituisce NULL come per qualsiasi altra stringa multilinea. Il comportamento precedente non era documentato, ma le persone che presumevano di avere i dati memorizzati come LINESTRING potrebbero trovare che questi ora restituiscono il valore NULL. |
postgis=# SELECT ST_AsText(ST_EndPoint('LINESTRING(1 1, 2 2, 3 3)'::geometry)); st_astext ------------ POINT(3 3) (1 row) postgis=# SELECT ST_EndPoint('POINT(1 1)'::geometry) IS NULL AS is_null; is_null ---------- t (1 row) --3d endpoint SELECT ST_AsEWKT(ST_EndPoint('LINESTRING(1 1 2, 1 2 3, 0 0 5)')); st_asewkt -------------- POINT(0 0 5) (1 row)
ST_Envelope — Returns a geometry representing the double precision (float8) bounding box of the supplied geometry.
geometry ST_Envelope(
geometry g1)
;
Returns the float8 minimum bounding box for the supplied geometry, as a geometry. The polygon is defined by the corner points of the bounding box ((MINX
, MINY
), (MINX
, MAXY
), (MAXX
, MAXY
), (MAXX
, MINY
), (MINX
, MINY
)). (PostGIS will add a ZMIN
/ZMAX
coordinate as well).
I casi degeneri (linee verticali, punti) restituiranno una geometria di dimensione inferiore al POLYGON
, cioè POINT
o LINESTRING
.
Disponibilità: il comportamento nella 1.5.0 è stato cambiato per dare in uscita numeri in precisione doppia anziche float4
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.1
This method implements the SQL/MM specification. SQL-MM 3: 5.1.15
SELECT ST_AsText(ST_Envelope('POINT(1 3)'::geometry)); st_astext ------------ POINT(1 3) (1 row) SELECT ST_AsText(ST_Envelope('LINESTRING(0 0, 1 3)'::geometry)); st_astext -------------------------------- POLYGON((0 0,0 3,1 3,1 0,0 0)) (1 row) SELECT ST_AsText(ST_Envelope('POLYGON((0 0, 0 1, 1.0000001 1, 1.0000001 0, 0 0))'::geometry)); st_astext -------------------------------------------------------------- POLYGON((0 0,0 1,1.00000011920929 1,1.00000011920929 0,0 0)) (1 row) SELECT ST_AsText(ST_Envelope('POLYGON((0 0, 0 1, 1.0000000001 1, 1.0000000001 0, 0 0))'::geometry)); st_astext -------------------------------------------------------------- POLYGON((0 0,0 1,1.00000011920929 1,1.00000011920929 0,0 0)) (1 row) SELECT Box3D(geom), Box2D(geom), ST_AsText(ST_Envelope(geom)) As envelopewkt FROM (SELECT 'POLYGON((0 0, 0 1000012333334.34545678, 1.0000001 1, 1.0000001 0, 0 0))'::geometry As geom) As foo;
ST_BoundingDiagonal — Returns the diagonal of the supplied geometry's bounding box.
geometry ST_BoundingDiagonal(
geometry geom, boolean fits=false)
;
Returns the diagonal of the supplied geometry's bounding box as linestring. If the input geometry is empty, the diagonal line is also empty, otherwise it is a 2-points linestring with minimum values of each dimension in its start point and maximum values in its end point.
The returned linestring geometry always retains SRID and dimensionality (Z and M presence) of the input geometry.
The fits
parameter specifies if the best fit is needed. If false, the diagonal of a somewhat larger bounding box can be accepted (is faster to obtain for geometries with a lot of vertices). In any case the bounding box of the returned diagonal line always covers the input geometry.
In degenerate cases (a single vertex in input) the returned linestring will be topologically invalid (no interior). This does not make the return semantically invalid. |
Availability: 2.2.0
This function supports 3d and will not drop the z-index.
This function supports M coordinates.
ST_ExteriorRing — Restituisce una LINESTRING che rappresenta l'anello esterno per una geometria POLYGON
. Restituisce NULL se la geometria non è un poligono. Non funziona con MULTIPOLYGON
geometry ST_ExteriorRing(
geometry a_polygon)
;
Restituisce una LINESTRING che rappresenta l'anello esterno per una geometria POLYGON
. Restituisce NULL se la geometria non è un poligono. Non funziona con MULTIPOLYGON
Funziona solamente con le geometrie di tipo POLYGON |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. 2.1.5.1
This method implements the SQL/MM specification. SQL-MM 3: 8.2.3, 8.3.3
This function supports 3d and will not drop the z-index.
--Se avete una tabella di poligoni: SELECT gid, ST_ExteriorRing(the_geom) AS ering FROM sometable; --Se avete una tabella MULTIPOLYGON --e volete riottenere un MULTILINESTRING composto dagli anelli esteriori mposed of the exterior rings of each polygon SELECT gid, ST_Collect(ST_ExteriorRing(the_geom)) AS erings FROM (SELECT gid, (ST_Dump(the_geom)).geom As the_geom FROM sometable) As foo GROUP BY gid; --Esempio 3d SELECT ST_AsEWKT( ST_ExteriorRing( ST_GeomFromEWKT('POLYGON((0 0 1, 1 1 1, 1 2 1, 1 1 1, 0 0 1))') ) ); st_asewkt --------- LINESTRING(0 0 1,1 1 1,1 2 1,1 1 1,0 0 1)
ST_GeometryN — Restituisce la geometria numero N (a partire da 1) se la geometria è una GEOMETRYCOLLECTION, (MULTI)POINT, (MULTI)LINESTRING, MULTICURVE o (MULTI)POLYGON, POLYHEDRALSURFACE. Altrimenti restituisce il valore NULL.
geometry ST_GeometryN(
geometry geomA, integer n)
;
Restituisce la geometria numero N (a partire da 1) se la geometria è una GEOMETRYCOLLECTION, (MULTI)POINT, (MULTI)LINESTRING, MULTICURVE o (MULTI)POLYGON, POLYHEDRALSURFACE. Altrimenti restituisce il valore NULL.
L'indice parte da 1, come per le specifiche OGC a partire dalla versione 0.8.0. Le precedenti versioni invece implementavano un indice a partire da zero. |
Se volete estrarre tutte le geometria, ST_Dump è più efficiente e funziona anche nel caso di geometrie singole. |
Migliorato: Nella 2.0.0 è stato introdotto il supporto per le superfici poliedriche, i triangoli e i TIN.
Changed: 2.0.0 Prior versions would return NULL for singular geometries. This was changed to return the geometry for ST_GeometryN(..,1) case.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 9.1.5
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
--Extracting a subset of points from a 3d multipoint SELECT n, ST_AsEWKT(ST_GeometryN(the_geom, n)) As geomewkt FROM ( VALUES (ST_GeomFromEWKT('MULTIPOINT(1 2 7, 3 4 7, 5 6 7, 8 9 10)') ), ( ST_GeomFromEWKT('MULTICURVE(CIRCULARSTRING(2.5 2.5,4.5 2.5, 3.5 3.5), (10 11, 12 11))') ) )As foo(the_geom) CROSS JOIN generate_series(1,100) n WHERE n <= ST_NumGeometries(the_geom); n | geomewkt ---+----------------------------------------- 1 | POINT(1 2 7) 2 | POINT(3 4 7) 3 | POINT(5 6 7) 4 | POINT(8 9 10) 1 | CIRCULARSTRING(2.5 2.5,4.5 2.5,3.5 3.5) 2 | LINESTRING(10 11,12 11) --Extracting all geometries (useful when you want to assign an id) SELECT gid, n, ST_GeometryN(the_geom, n) FROM sometable CROSS JOIN generate_series(1,100) n WHERE n <= ST_NumGeometries(the_geom);
-- Esempio con superficie poliedrica -- Spezza una superficie poliedrica nelle facce che la compongono SELECT ST_AsEWKT(ST_GeometryN(p_geom,3)) As geom_ewkt FROM (SELECT ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )') AS p_geom ) AS a; geom_ewkt ------------------------------------------ POLYGON((0 0 0,1 0 0,1 0 1,0 0 1,0 0 0))
-- TIN -- SELECT ST_AsEWKT(ST_GeometryN(geom,2)) as wkt FROM (SELECT ST_GeomFromEWKT('TIN ((( 0 0 0, 0 0 1, 0 1 0, 0 0 0 )), (( 0 0 0, 0 1 0, 1 1 0, 0 0 0 )) )') AS geom ) AS g; -- result -- wkt ------------------------------------- TRIANGLE((0 0 0,0 1 0,1 1 0,0 0 0))
ST_GeometryType — Restituisce il tipo di geometria per il valore ST_Geometry.
text ST_GeometryType(
geometry g1)
;
Restituisce il tipo di geometria come stringa. P. es.: 'ST_Linestring', 'ST_Polygon','ST_MultiPolygon' ecc. Questa funzione differisce da GeometryType(geometry) per il prefisso ST che viene restituito, così come per il fatto che non indica se la geometria è misurata.
Miglioramento: nella versione 2.0.0 è stato introdotto il supporto per le superfici poliedriche.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.4
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
SELECT ST_GeometryType(ST_GeomFromText('LINESTRING(77.29 29.07,77.42 29.26,77.27 29.31,77.29 29.07)')); --risultato ST_LineString
SELECT ST_GeometryType(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')); --risultato ST_PolyhedralSurface
SELECT ST_GeometryType(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')); --risultato ST_PolyhedralSurface
SELECT ST_GeometryType(geom) as result FROM (SELECT ST_GeomFromEWKT('TIN ((( 0 0 0, 0 0 1, 0 1 0, 0 0 0 )), (( 0 0 0, 0 1 0, 1 1 0, 0 0 0 )) )') AS geom ) AS g; result -------- ST_Tin
ST_InteriorRingN — Return the Nth interior linestring ring of the polygon geometry. Return NULL if the geometry is not a polygon or the given N is out of range.
geometry ST_InteriorRingN(
geometry a_polygon, integer n)
;
Return the Nth interior linestring ring of the polygon geometry. Return NULL if the geometry is not a polygon or the given N is out of range. index starts at 1.
Non funzionerà per i MULTIPOLYGON. Da utilizzare assieme a ST_Dump per i MULTIPOLYGON |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 8.2.6, 8.3.5
This function supports 3d and will not drop the z-index.
SELECT ST_AsText(ST_InteriorRingN(the_geom, 1)) As the_geom FROM (SELECT ST_BuildArea( ST_Collect(ST_Buffer(ST_Point(1,2), 20,3), ST_Buffer(ST_Point(1, 2), 10,3))) As the_geom ) as foo
ST_ExteriorRing ST_BuildArea, ST_Collect, ST_Dump, ST_NumInteriorRing, ST_NumInteriorRings
ST_IsClosed — Restituisce TRUE
se il punto iniziale e quello finale di LINESTRING
coincidono. Per le superfici poliedriche indica una superficie chiusa (volumetrica).
boolean ST_IsClosed(
geometry g)
;
Restituisce TRUE
se il punto iniziale e quello finale di LINESTRING
coincidono. Per le superfici poliedriche indicase la superficie è un'area (aperta) o un volume (chiusa).
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 7.1.5, 9.3.3
SQL-MM definisce il risultato di |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
Miglioramento: nella versione 2.0.0 è stato introdotto il supporto per le superfici poliedriche.
This function supports Polyhedral surfaces.
postgis=# SELECT ST_IsClosed('LINESTRING(0 0, 1 1)'::geometry); st_isclosed ------------- f (1 row) postgis=# SELECT ST_IsClosed('LINESTRING(0 0, 0 1, 1 1, 0 0)'::geometry); st_isclosed ------------- t (1 row) postgis=# SELECT ST_IsClosed('MULTILINESTRING((0 0, 0 1, 1 1, 0 0),(0 0, 1 1))'::geometry); st_isclosed ------------- f (1 row) postgis=# SELECT ST_IsClosed('POINT(0 0)'::geometry); st_isclosed ------------- t (1 row) postgis=# SELECT ST_IsClosed('MULTIPOINT((0 0), (1 1))'::geometry); st_isclosed ------------- t (1 row)
-- Un cubo -- SELECT ST_IsClosed(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')); st_isclosed ------------- t -- Come un cubo, ma senza un lato -- SELECT ST_IsClosed(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)) )')); st_isclosed ------------- f
ST_IsCollection — Restituisce TRUE
se l'argomento è una collezione (MULTI*
, GEOMETRYCOLLECTION
, ...)
boolean ST_IsCollection(
geometry g)
;
Restituisce TRUE
se il tipo di geometria è uno tra:
GEOMETRYCOLLECTION
MULTI{POINT,POLYGON,LINESTRING,CURVE,SURFACE}
COMPOUNDCURVE
Questa funzione analizza il tipo di geometria. Significa che restituirà |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
postgis=# SELECT ST_IsCollection('LINESTRING(0 0, 1 1)'::geometry); st_iscollection ------------- f (1 row) postgis=# SELECT ST_IsCollection('MULTIPOINT EMPTY'::geometry); st_iscollection ------------- t (1 row) postgis=# SELECT ST_IsCollection('MULTIPOINT((0 0))'::geometry); st_iscollection ------------- t (1 row) postgis=# SELECT ST_IsCollection('MULTIPOINT((0 0), (42 42))'::geometry); st_iscollection ------------- t (1 row) postgis=# SELECT ST_IsCollection('GEOMETRYCOLLECTION(POINT(0 0))'::geometry); st_iscollection ------------- t (1 row)
ST_IsEmpty — Restituisce TRUE se la geometria è una geometrycollection, un poligono o un punto vuoto, ecc.
boolean ST_IsEmpty(
geometry geomA)
;
Restituisce TRUE se la Geometry è una geometria vuota. Se è TRUE, allora questa Geometry rappresenta una geometria vuota (una collezione, un poligono, un punto, ecc.)
SQL-MM definisce il risultato di ST_IsEmpty(NULL) come 0, mentre PostGIS restituisce NULL. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.1
This method implements the SQL/MM specification. SQL-MM 3: 5.1.7
This method supports Circular Strings and Curves
Changed: 2.0.0 In prior versions of PostGIS ST_GeomFromText('GEOMETRYCOLLECTION(EMPTY)') was allowed. This is now illegal in PostGIS 2.0.0 to better conform with SQL/MM standards |
SELECT ST_IsEmpty(ST_GeomFromText('GEOMETRYCOLLECTION EMPTY')); st_isempty ------------ t (1 row) SELECT ST_IsEmpty(ST_GeomFromText('POLYGON EMPTY')); st_isempty ------------ t (1 row) SELECT ST_IsEmpty(ST_GeomFromText('POLYGON((1 2, 3 4, 5 6, 1 2))')); st_isempty ------------ f (1 row) SELECT ST_IsEmpty(ST_GeomFromText('POLYGON((1 2, 3 4, 5 6, 1 2))')) = false; ?column? ---------- t (1 row) SELECT ST_IsEmpty(ST_GeomFromText('CIRCULARSTRING EMPTY')); st_isempty ------------ t (1 row)
ST_IsRing — Returns TRUE
if this LINESTRING
is both closed and simple.
boolean ST_IsRing(
geometry g)
;
Restituisce TRUE
se questa LINESTRING
è sia ST_IsClosed (ST_StartPoint(
g
)~=
ST_Endpoint(
) che ST_IsSimple (non interseca se stessa).g
)
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. 2.1.5.1
This method implements the SQL/MM specification. SQL-MM 3: 7.1.6
SQL-MM defines the result of |
SELECT ST_IsRing(the_geom), ST_IsClosed(the_geom), ST_IsSimple(the_geom) FROM (SELECT 'LINESTRING(0 0, 0 1, 1 1, 1 0, 0 0)'::geometry AS the_geom) AS foo; st_isring | st_isclosed | st_issimple -----------+-------------+------------- t | t | t (1 row) SELECT ST_IsRing(the_geom), ST_IsClosed(the_geom), ST_IsSimple(the_geom) FROM (SELECT 'LINESTRING(0 0, 0 1, 1 0, 1 1, 0 0)'::geometry AS the_geom) AS foo; st_isring | st_isclosed | st_issimple -----------+-------------+------------- f | t | f (1 row)
ST_IsSimple — Returns (TRUE) if this Geometry has no anomalous geometric points, such as self intersection or self tangency.
boolean ST_IsSimple(
geometry geomA)
;
Returns true if this Geometry has no anomalous geometric points, such as self intersection or self tangency. For more information on the OGC's definition of geometry simplicity and validity, refer to "Ensuring OpenGIS compliancy of geometries"
SQL-MM defines the result of ST_IsSimple(NULL) to be 0, while PostGIS returns NULL. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.1
This method implements the SQL/MM specification. SQL-MM 3: 5.1.8
This function supports 3d and will not drop the z-index.
ST_IsValid — Returns true
if the ST_Geometry
is well formed.
boolean ST_IsValid(
geometry g)
;
boolean ST_IsValid(
geometry g, integer flags)
;
Test if an ST_Geometry value is well formed. For geometries that are invalid, the PostgreSQL NOTICE will provide details of why it is not valid. For more information on the OGC's definition of geometry simplicity and validity, refer to "Ensuring OpenGIS compliancy of geometries"
SQL-MM definisce il risultato di |
The version accepting flags is available starting with 2.0.0 and requires GEOS >= 3.3.0. Such version does not print a NOTICE explaining the invalidity. Allowed flags
are documented in ST_IsValidDetail.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.9
ST_IsValidReason — Returns text stating if a geometry is valid or not and if not valid, a reason why.
text ST_IsValidReason(
geometry geomA)
;
text ST_IsValidReason(
geometry geomA, integer flags)
;
Returns text stating if a geometry is valid or not an if not valid, a reason why.
Useful in combination with ST_IsValid to generate a detailed report of invalid geometries and reasons.
Allowed flags
are documented in ST_IsValidDetail.
Availability: 1.4 - requires GEOS >= 3.1.0.
Availability: 2.0 - requires GEOS >= 3.3.0 for the version taking flags.
--First 3 Rejects from a successful quintuplet experiment SELECT gid, ST_IsValidReason(the_geom) as validity_info FROM (SELECT ST_MakePolygon(ST_ExteriorRing(e.buff), ST_Accum(f.line)) As the_geom, gid FROM (SELECT ST_Buffer(ST_MakePoint(x1*10,y1), z1) As buff, x1*10 + y1*100 + z1*1000 As gid FROM generate_series(-4,6) x1 CROSS JOIN generate_series(2,5) y1 CROSS JOIN generate_series(1,8) z1 WHERE x1 > y1*0.5 AND z1 < x1*y1) As e INNER JOIN (SELECT ST_Translate(ST_ExteriorRing(ST_Buffer(ST_MakePoint(x1*10,y1), z1)),y1*1, z1*2) As line FROM generate_series(-3,6) x1 CROSS JOIN generate_series(2,5) y1 CROSS JOIN generate_series(1,10) z1 WHERE x1 > y1*0.75 AND z1 < x1*y1) As f ON (ST_Area(e.buff) > 78 AND ST_Contains(e.buff, f.line)) GROUP BY gid, e.buff) As quintuplet_experiment WHERE ST_IsValid(the_geom) = false ORDER BY gid LIMIT 3; gid | validity_info ------+-------------------------- 5330 | Self-intersection [32 5] 5340 | Self-intersection [42 5] 5350 | Self-intersection [52 5] --simple example SELECT ST_IsValidReason('LINESTRING(220227 150406,2220227 150407,222020 150410)'); st_isvalidreason ------------------ Valid Geometry
ST_IsValidDetail — Returns a valid_detail (valid,reason,location) row stating if a geometry is valid or not and if not valid, a reason why and a location where.
valid_detail ST_IsValidDetail(
geometry geom)
;
valid_detail ST_IsValidDetail(
geometry geom, integer flags)
;
Returns a valid_detail row, formed by a boolean (valid) stating if a geometry is valid, a varchar (reason) stating a reason why it is invalid and a geometry (location) pointing out where it is invalid.
Useful to substitute and improve the combination of ST_IsValid and ST_IsValidReason to generate a detailed report of invalid geometries.
The 'flags' argument is a bitfield. It can have the following values:
1: Consider self-intersecting rings forming holes as valid. This is also know as "the ESRI flag". Note that this is against the OGC model.
Availability: 2.0.0 - requires GEOS >= 3.3.0.
--First 3 Rejects from a successful quintuplet experiment SELECT gid, reason(ST_IsValidDetail(the_geom)), ST_AsText(location(ST_IsValidDetail(the_geom))) as location FROM (SELECT ST_MakePolygon(ST_ExteriorRing(e.buff), ST_Accum(f.line)) As the_geom, gid FROM (SELECT ST_Buffer(ST_MakePoint(x1*10,y1), z1) As buff, x1*10 + y1*100 + z1*1000 As gid FROM generate_series(-4,6) x1 CROSS JOIN generate_series(2,5) y1 CROSS JOIN generate_series(1,8) z1 WHERE x1 > y1*0.5 AND z1 < x1*y1) As e INNER JOIN (SELECT ST_Translate(ST_ExteriorRing(ST_Buffer(ST_MakePoint(x1*10,y1), z1)),y1*1, z1*2) As line FROM generate_series(-3,6) x1 CROSS JOIN generate_series(2,5) y1 CROSS JOIN generate_series(1,10) z1 WHERE x1 > y1*0.75 AND z1 < x1*y1) As f ON (ST_Area(e.buff) > 78 AND ST_Contains(e.buff, f.line)) GROUP BY gid, e.buff) As quintuplet_experiment WHERE ST_IsValid(the_geom) = false ORDER BY gid LIMIT 3; gid | reason | location ------+-------------------+------------- 5330 | Self-intersection | POINT(32 5) 5340 | Self-intersection | POINT(42 5) 5350 | Self-intersection | POINT(52 5) --simple example SELECT * FROM ST_IsValidDetail('LINESTRING(220227 150406,2220227 150407,222020 150410)'); valid | reason | location -------+--------+---------- t | |
ST_M — Return the M coordinate of the point, or NULL if not available. Input must be a point.
float ST_M(
geometry a_point)
;
Return the M coordinate of the point, or NULL if not available. Input must be a point.
This is not (yet) part of the OGC spec, but is listed here to complete the point coordinate extractor function list. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification.
This function supports 3d and will not drop the z-index.
ST_NDims — Returns coordinate dimension of the geometry as a small int. Values are: 2,3 or 4.
integer ST_NDims(
geometry g1)
;
Returns the coordinate dimension of the geometry. PostGIS supports 2 - (x,y) , 3 - (x,y,z) or 2D with measure - x,y,m, and 4 - 3D with measure space x,y,z,m
This function supports 3d and will not drop the z-index.
ST_NPoints — Return the number of points (vertexes) in a geometry.
integer ST_NPoints(
geometry g1)
;
Return the number of points in a geometry. Works for all geometries.
Miglioramento: nella versione 2.0.0 è stato introdotto il supporto per le superfici poliedriche.
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
ST_NRings — If the geometry is a polygon or multi-polygon returns the number of rings.
integer ST_NRings(
geometry geomA)
;
If the geometry is a polygon or multi-polygon returns the number of rings. Unlike NumInteriorRings, it counts the outer rings as well.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
ST_NumGeometries — If geometry is a GEOMETRYCOLLECTION (or MULTI*) return the number of geometries, for single geometries will return 1, otherwise return NULL.
integer ST_NumGeometries(
geometry geom)
;
Returns the number of Geometries. If geometry is a GEOMETRYCOLLECTION (or MULTI*) return the number of geometries, for single geometries will return 1, otherwise return NULL.
Migliorato: Nella 2.0.0 è stato introdotto il supporto per le superfici poliedriche, i triangoli e i TIN.
Changed: 2.0.0 In prior versions this would return NULL if the geometry was not a collection/MULTI type. 2.0.0+ now returns 1 for single geometries e.g POLYGON, LINESTRING, POINT.
This method implements the SQL/MM specification. SQL-MM 3: 9.1.4
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
--Prior versions would have returned NULL for this -- in 2.0.0 this returns 1 SELECT ST_NumGeometries(ST_GeomFromText('LINESTRING(77.29 29.07,77.42 29.26,77.27 29.31,77.29 29.07)')); --result 1 --Geometry Collection Example - multis count as one geom in a collection SELECT ST_NumGeometries(ST_GeomFromEWKT('GEOMETRYCOLLECTION(MULTIPOINT(-2 3 , -2 2), LINESTRING(5 5 ,10 10), POLYGON((-7 4.2,-7.1 5,-7.1 4.3,-7 4.2)))')); --result 3
ST_NumInteriorRings — Return the number of interior rings of the a polygon in the geometry. This will work with POLYGON and return NULL for a MULTIPOLYGON type or any other type
integer ST_NumInteriorRings(
geometry a_polygon)
;
Return the number of interior rings of the first polygon in the geometry. This will work with both POLYGON and MULTIPOLYGON types but only looks at the first polygon. Return NULL if there is no polygon in the geometry.
This method implements the SQL/MM specification. SQL-MM 3: 8.2.5
Changed: 2.0.0 - in prior versions it would return the number of interior rings for the first POLYGON in a MULTIPOLYGON.
--If you have a regular polygon SELECT gid, field1, field2, ST_NumInteriorRings(the_geom) AS numholes FROM sometable; --If you have multipolygons --And you want to know the total number of interior rings in the MULTIPOLYGON SELECT gid, field1, field2, SUM(ST_NumInteriorRings(the_geom)) AS numholes FROM (SELECT gid, field1, field2, (ST_Dump(the_geom)).geom As the_geom FROM sometable) As foo GROUP BY gid, field1,field2;
ST_NumInteriorRing — Return the number of interior rings of a polygon in the geometry. Synonym for ST_NumInteriorRings.
integer ST_NumInteriorRing(
geometry a_polygon)
;
ST_NumPatches — Return the number of faces on a Polyhedral Surface. Will return null for non-polyhedral geometries.
integer ST_NumPatches(
geometry g1)
;
Return the number of faces on a Polyhedral Surface. Will return null for non-polyhedral geometries. This is an alias for ST_NumGeometries to support MM naming. Faster to use ST_NumGeometries if you don't care about MM convention.
Availability: 2.0.0
This function supports 3d and will not drop the z-index.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: ?
This function supports Polyhedral surfaces.
SELECT ST_NumPatches(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')); --risultato 6
ST_NumPoints — Return the number of points in an ST_LineString or ST_CircularString value.
integer ST_NumPoints(
geometry g1)
;
Return the number of points in an ST_LineString or ST_CircularString value. Prior to 1.4 only works with Linestrings as the specs state. From 1.4 forward this is an alias for ST_NPoints which returns number of vertexes for not just line strings. Consider using ST_NPoints instead which is multi-purpose and works with many geometry types.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 7.2.4
ST_PatchN — Return the 1-based Nth geometry (face) if the geometry is a POLYHEDRALSURFACE, POLYHEDRALSURFACEM. Otherwise, return NULL.
geometry ST_PatchN(
geometry geomA, integer n)
;
>Return the 1-based Nth geometry (face) if the geometry is a POLYHEDRALSURFACE, POLYHEDRALSURFACEM. Otherwise, return NULL. This returns the same answer as ST_GeometryN for Polyhedral Surfaces. Using ST_GemoetryN is faster.
L'indice parte da 1. |
If you want to extract all geometries, of a geometry, ST_Dump is more efficient. |
Availability: 2.0.0
This method implements the SQL/MM specification. SQL-MM 3: ?
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
--Extract the 2nd face of the polyhedral surface SELECT ST_AsEWKT(ST_PatchN(geom, 2)) As geomewkt FROM ( VALUES (ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')) ) As foo(geom); geomewkt ---+----------------------------------------- POLYGON((0 0 0,0 1 0,1 1 0,1 0 0,0 0 0))
ST_PointN — Return the Nth point in the first linestring or circular linestring in the geometry. Return NULL if there is no linestring in the geometry.
geometry ST_PointN(
geometry a_linestring, integer n)
;
Return the Nth point in a single linestring or circular linestring in the geometry. Return NULL if there is no linestring in the geometry.
L'indice parte da 1, come per le specifiche OGC a partire dalla versione 0.8.0. Le precedenti versioni invece implementavano un indice a partire da zero. |
If you want to get the nth point of each line string in a multilinestring, use in conjunction with ST_Dump |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 7.2.5, 7.3.5
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
Changed: 2.0.0 no longer works with single geometry multilinestrings. In older versions of PostGIS -- a single line multilinestring would work happily with this function and return the start point. In 2.0.0 it just returns NULL like any other multilinestring. |
-- Estrarre tutti i POINT da una LINESTRING SELECT ST_AsText( ST_PointN( column1, generate_series(1, ST_NPoints(column1)) )) FROM ( VALUES ('LINESTRING(0 0, 1 1, 2 2)'::geometry) ) AS foo; st_astext ------------ POINT(0 0) POINT(1 1) POINT(2 2) (3 rows) --Esempio di stringa circolare SELECT ST_AsText(ST_PointN(ST_GeomFromText('CIRCULARSTRING(1 2, 3 2, 1 2)'),2)); st_astext ---------- POINT(3 2)
ST_NPoints — Returns a MultiPoint containing all of the coordinates of a geometry.
geometry ST_StartPoint(
geometry geomA)
;
Returns a MultiPoint containing all of the coordinates of a geometry. Does not remove points that are duplicated in the input geometry, including start and end points of ring geometries. (If this behavior is undesired, duplicates may be removed using ST_RemoveRepeatedPoints).
M and Z ordinates will be preserved if present.
This method supports Circular Strings and Curves
This function supports 3d and will not drop the z-index.
ST_SRID — Returns the spatial reference identifier for the ST_Geometry as defined in spatial_ref_sys table.
integer ST_SRID(
geometry g1)
;
Returns the spatial reference identifier for the ST_Geometry as defined in spatial_ref_sys table. Section 4.3.1, “The SPATIAL_REF_SYS Table and Spatial Reference Systems”
spatial_ref_sys table is a table that catalogs all spatial reference systems known to PostGIS and is used for transformations from one spatial reference system to another. So verifying you have the right spatial reference system identifier is important if you plan to ever transform your geometries. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.1
This method implements the SQL/MM specification. SQL-MM 3: 5.1.5
This method supports Circular Strings and Curves
ST_StartPoint — Returns the first point of a LINESTRING
geometry as a POINT
.
geometry ST_StartPoint(
geometry geomA)
;
Returns the first point of a LINESTRING
or CIRCULARLINESTRING
geometry as a POINT
or NULL
if the input parameter is not a LINESTRING
or CIRCULARLINESTRING
.
This method implements the SQL/MM specification. SQL-MM 3: 7.1.3
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
Modifica: La versione 2.0.0 non funziona più con geometrie singole di stringhe multilinea. Nelle versioni precedenti di PostGIS una stringa multilinea con una sola linea avrebbe funzionato tranquillamente con questa funzione, restituendo il punto di inizio. Nella versione 2.0.0 la funzione restituisce NULL come per qualsiasi altra stringa multilinea. Il comportamento precedente non era documentato, ma le persone che presumevano di avere i dati memorizzati come LINESTRING potrebbero trovare che questi ora restituiscono il valore NULL. |
SELECT ST_AsText(ST_StartPoint('LINESTRING(0 1, 0 2)'::geometry)); st_astext ------------ POINT(0 1) (1 row) SELECT ST_StartPoint('POINT(0 1)'::geometry) IS NULL AS is_null; is_null ---------- t (1 row) --3d line SELECT ST_AsEWKT(ST_StartPoint('LINESTRING(0 1 1, 0 2 2)'::geometry)); st_asewkt ------------ POINT(0 1 1) (1 row) -- circular linestring -- SELECT ST_AsText(ST_StartPoint('CIRCULARSTRING(5 2,-3 1.999999, -2 1, -4 2, 5 2)'::geometry)); st_astext ------------ POINT(5 2)
ST_Summary — Returns a text summary of the contents of the geometry.
text ST_Summary(
geometry g)
;
text ST_Summary(
geography g)
;
Returns a text summary of the contents of the geometry.
Flags shown square brackets after the geometry type have the following meaning:
M: has M ordinate
Z: has Z ordinate
B: has a cached bounding box
G: is geodetic (geography)
S: has spatial reference system
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
Availability: 1.2.2
Enhanced: 2.0.0 added support for geography
Enhanced: 2.1.0 S flag to denote if has a known spatial reference system
Enhanced: 2.2.0 Added support for TIN and Curves
=# SELECT ST_Summary(ST_GeomFromText('LINESTRING(0 0, 1 1)')) as geom, ST_Summary(ST_GeogFromText('POLYGON((0 0, 1 1, 1 2, 1 1, 0 0))')) geog; geom | geog -----------------------------+-------------------------- LineString[B] with 2 points | Polygon[BGS] with 1 rings | ring 0 has 5 points : (1 row) =# SELECT ST_Summary(ST_GeogFromText('LINESTRING(0 0 1, 1 1 1)')) As geog_line, ST_Summary(ST_GeomFromText('SRID=4326;POLYGON((0 0 1, 1 1 2, 1 2 3, 1 1 1, 0 0 1))')) As geom_poly; ; geog_line | geom_poly -------------------------------- +-------------------------- LineString[ZBGS] with 2 points | Polygon[ZBS] with 1 rings : ring 0 has 5 points : (1 row)
ST_X — Restituisce la coordinata X del punto, o NULL se non disponibile. L'input deve essere un punto.
float ST_X(
geometry a_point)
;
Restituisce la coordinata X del punto, o NULL se non disponibile. L'input deve essere un punto.
Se volete recuperare i valori minimo e massimo della x per una qualsiasi geometria, considerate le funzioni ST_XMin, ST_XMax. |
This method implements the SQL/MM specification. SQL-MM 3: 6.1.3
This function supports 3d and will not drop the z-index.
ST_XMax — Returns X maxima of a bounding box 2d or 3d or a geometry.
float ST_XMax(
box3d aGeomorBox2DorBox3D)
;
Returns X maxima of a bounding box 2d or 3d or a geometry.
Although this function is only defined for box3d, it will work for box2d and geometry because of the auto-casting behavior defined for geometries and box2d. However you can not feed it a geometry or box2d text representation, since that will not auto-cast. |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_XMax('BOX3D(1 2 3, 4 5 6)'); st_xmax ------- 4 SELECT ST_XMax(ST_GeomFromText('LINESTRING(1 3 4, 5 6 7)')); st_xmax ------- 5 SELECT ST_XMax(CAST('BOX(-3 2, 3 4)' As box2d)); st_xmax ------- 3 --NOTA: QUESTO NON FUNZIONA perché cercherà di eseguire il casting automatico della rappresentazione della stringa a un BOX3D SELECT ST_XMax('LINESTRING(1 3, 5 6)'); --ERROR: BOX3D parser - doesn't start with BOX3D( SELECT ST_XMax(ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)')); st_xmax -------- 220288.248780547
ST_XMin — Returns X minima of a bounding box 2d or 3d or a geometry.
float ST_XMin(
box3d aGeomorBox2DorBox3D)
;
Returns X minima of a bounding box 2d or 3d or a geometry.
Although this function is only defined for box3d, it will work for box2d and geometry because of the auto-casting behavior defined for geometries and box2d. However you can not feed it a geometry or box2d text representation, since that will not auto-cast. |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_XMin('BOX3D(1 2 3, 4 5 6)'); st_xmin ------- 1 SELECT ST_XMin(ST_GeomFromText('LINESTRING(1 3 4, 5 6 7)')); st_xmin ------- 1 SELECT ST_XMin(CAST('BOX(-3 2, 3 4)' As box2d)); st_xmin ------- -3 --NOTA: QUESTO NON FUNZIONA perché cercherà di eseguire il casting automatico della rappresentazione della stringa a un BOX3D SELECT ST_XMin('LINESTRING(1 3, 5 6)'); --ERROR: BOX3D parser - doesn't start with BOX3D( SELECT ST_XMin(ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)')); st_xmin -------- 220186.995121892
ST_Y — Return the Y coordinate of the point, or NULL if not available. Input must be a point.
float ST_Y(
geometry a_point)
;
Return the Y coordinate of the point, or NULL if not available. Input must be a point.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 6.1.4
This function supports 3d and will not drop the z-index.
ST_YMax — Returns Y maxima of a bounding box 2d or 3d or a geometry.
float ST_YMax(
box3d aGeomorBox2DorBox3D)
;
Returns Y maxima of a bounding box 2d or 3d or a geometry.
Although this function is only defined for box3d, it will work for box2d and geometry because of the auto-casting behavior defined for geometries and box2d. However you can not feed it a geometry or box2d text representation, since that will not auto-cast. |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_YMax('BOX3D(1 2 3, 4 5 6)'); st_ymax ------- 5 SELECT ST_YMax(ST_GeomFromText('LINESTRING(1 3 4, 5 6 7)')); st_ymax ------- 6 SELECT ST_YMax(CAST('BOX(-3 2, 3 4)' As box2d)); st_ymax ------- 4 --NOTA: QUESTO NON FUNZIONA perché cercherà di eseguire il casting automatico della rappresentazione della stringa a un BOX3D SELECT ST_YMax('LINESTRING(1 3, 5 6)'); --ERROR: BOX3D parser - doesn't start with BOX3D( SELECT ST_YMax(ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)')); st_ymax -------- 150506.126829327
ST_YMin — Returns Y minima of a bounding box 2d or 3d or a geometry.
float ST_YMin(
box3d aGeomorBox2DorBox3D)
;
Returns Y minima of a bounding box 2d or 3d or a geometry.
Although this function is only defined for box3d, it will work for box2d and geometry because of the auto-casting behavior defined for geometries and box2d. However you can not feed it a geometry or box2d text representation, since that will not auto-cast. |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_YMin('BOX3D(1 2 3, 4 5 6)'); st_ymin ------- 2 SELECT ST_YMin(ST_GeomFromText('LINESTRING(1 3 4, 5 6 7)')); st_ymin ------- 3 SELECT ST_YMin(CAST('BOX(-3 2, 3 4)' As box2d)); st_ymin ------- 2 --Observe THIS DOES NOT WORK because it will try to autocast the string representation to a BOX3D SELECT ST_YMin('LINESTRING(1 3, 5 6)'); --ERROR: BOX3D parser - doesn't start with BOX3D( SELECT ST_YMin(ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)')); st_ymin -------- 150406
ST_Z — Return the Z coordinate of the point, or NULL if not available. Input must be a point.
float ST_Z(
geometry a_point)
;
ST_ZMax — Returns Z minima of a bounding box 2d or 3d or a geometry.
float ST_ZMax(
box3d aGeomorBox2DorBox3D)
;
Returns Z maxima of a bounding box 2d or 3d or a geometry.
Although this function is only defined for box3d, it will work for box2d and geometry because of the auto-casting behavior defined for geometries and box2d. However you can not feed it a geometry or box2d text representation, since that will not auto-cast. |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_ZMax('BOX3D(1 2 3, 4 5 6)'); st_zmax ------- 6 SELECT ST_ZMax(ST_GeomFromEWKT('LINESTRING(1 3 4, 5 6 7)')); st_zmax ------- 7 SELECT ST_ZMax('BOX3D(-3 2 1, 3 4 1)' ); st_zmax ------- 1 --Observe THIS DOES NOT WORK because it will try to autocast the string representation to a BOX3D SELECT ST_ZMax('LINESTRING(1 3 4, 5 6 7)'); --ERROR: BOX3D parser - doesn't start with BOX3D( SELECT ST_ZMax(ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)')); st_zmax -------- 3
ST_Zmflag — Returns ZM (dimension semantic) flag of the geometries as a small int. Values are: 0=2d, 1=3dm, 2=3dz, 3=4d.
smallint ST_Zmflag(
geometry geomA)
;
Returns ZM (dimension semantic) flag of the geometries as a small int. Values are: 0=2d, 1=3dm, 2=3dz, 3=4d.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_Zmflag(ST_GeomFromEWKT('LINESTRING(1 2, 3 4)')); st_zmflag ----------- 0 SELECT ST_Zmflag(ST_GeomFromEWKT('LINESTRINGM(1 2 3, 3 4 3)')); st_zmflag ----------- 1 SELECT ST_Zmflag(ST_GeomFromEWKT('CIRCULARSTRING(1 2 3, 3 4 3, 5 6 3)')); st_zmflag ----------- 2 SELECT ST_Zmflag(ST_GeomFromEWKT('POINT(1 2 3 4)')); st_zmflag ----------- 3
ST_ZMin — Returns Z minima of a bounding box 2d or 3d or a geometry.
float ST_ZMin(
box3d aGeomorBox2DorBox3D)
;
Restituisce la minima Z per una bounding box 2d o 3d o per una geometria.
Although this function is only defined for box3d, it will work for box2d and geometry because of the auto-casting behavior defined for geometries and box2d. However you can not feed it a geometry or box2d text representation, since that will not auto-cast. |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_ZMin('BOX3D(1 2 3, 4 5 6)'); st_zmin ------- 3 SELECT ST_ZMin(ST_GeomFromEWKT('LINESTRING(1 3 4, 5 6 7)')); st_zmin ------- 4 SELECT ST_ZMin('BOX3D(-3 2 1, 3 4 1)' ); st_zmin ------- 1 --NOTA: QUESTO NON FUNZIONA perché cercherà di eseguire il casting automatico della rappresentazione della stringa a un BOX3D SELECT ST_ZMin('LINESTRING(1 3 4, 5 6 7)'); --ERROR: BOX3D parser - doesn't start with BOX3D( SELECT ST_ZMin(ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)')); st_zmin -------- 1
ST_AddPoint — Adds a point to a LineString before point <position> (0-based index).
geometry ST_AddPoint(
geometry linestring, geometry point)
;
geometry ST_AddPoint(
geometry linestring, geometry point, integer position)
;
Adds a point to a LineString before point <position> (0-based index). Third parameter can be omitted or set to -1 for appending.
Availability: 1.1.0
This function supports 3d and will not drop the z-index.
--guarantee all linestrings in a table are closed --by adding the start point of each linestring to the end of the line string --only for those that are not closed UPDATE sometable SET the_geom = ST_AddPoint(the_geom, ST_StartPoint(the_geom)) FROM sometable WHERE ST_IsClosed(the_geom) = false; --Adding point to a 3-d line SELECT ST_AsEWKT(ST_AddPoint(ST_GeomFromEWKT('LINESTRING(0 0 1, 1 1 1)'), ST_MakePoint(1, 2, 3))); --result st_asewkt ---------- LINESTRING(0 0 1,1 1 1,1 2 3)
ST_Affine — Applies a 3d affine transformation to the geometry to do things like translate, rotate, scale in one step.
geometry ST_Affine(
geometry geomA, float a, float b, float c, float d, float e, float f, float g, float h, float i, float xoff, float yoff, float zoff)
;
geometry ST_Affine(
geometry geomA, float a, float b, float d, float e, float xoff, float yoff)
;
Applies a 3d affine transformation to the geometry to do things like translate, rotate, scale in one step.
Version 1: The call
ST_Affine(geom, a, b, c, d, e, f, g, h, i, xoff, yoff, zoff)
represents the transformation matrix
/ a b c xoff \ | d e f yoff | | g h i zoff | \ 0 0 0 1 /
and the vertices are transformed as follows:
x' = a*x + b*y + c*z + xoff y' = d*x + e*y + f*z + yoff z' = g*x + h*y + i*z + zoff
All of the translate / scale functions below are expressed via such an affine transformation.
Version 2: Applies a 2d affine transformation to the geometry. The call
ST_Affine(geom, a, b, d, e, xoff, yoff)
represents the transformation matrix
/ a b 0 xoff \ / a b xoff \ | d e 0 yoff | rsp. | d e yoff | | 0 0 1 0 | \ 0 0 1 / \ 0 0 0 1 /
and the vertices are transformed as follows:
x' = a*x + b*y + xoff y' = d*x + e*y + yoff z' = z
This method is a subcase of the 3D method above.
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Availability: 1.1.2. Name changed from Affine to ST_Affine in 1.2.2
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
--Rotate a 3d line 180 degrees about the z axis. Note this is long-hand for doing ST_Rotate(); SELECT ST_AsEWKT(ST_Affine(the_geom, cos(pi()), -sin(pi()), 0, sin(pi()), cos(pi()), 0, 0, 0, 1, 0, 0, 0)) As using_affine, ST_AsEWKT(ST_Rotate(the_geom, pi())) As using_rotate FROM (SELECT ST_GeomFromEWKT('LINESTRING(1 2 3, 1 4 3)') As the_geom) As foo; using_affine | using_rotate -----------------------------+----------------------------- LINESTRING(-1 -2 3,-1 -4 3) | LINESTRING(-1 -2 3,-1 -4 3) (1 row) --Rotate a 3d line 180 degrees in both the x and z axis SELECT ST_AsEWKT(ST_Affine(the_geom, cos(pi()), -sin(pi()), 0, sin(pi()), cos(pi()), -sin(pi()), 0, sin(pi()), cos(pi()), 0, 0, 0)) FROM (SELECT ST_GeomFromEWKT('LINESTRING(1 2 3, 1 4 3)') As the_geom) As foo; st_asewkt ------------------------------- LINESTRING(-1 -2 -3,-1 -4 -3) (1 row)
ST_Force2D — Forces the geometries into a "2-dimensional mode" so that all output representations will only have the X and Y coordinates.
geometry ST_Force2D(
geometry geomA)
;
Forces the geometries into a "2-dimensional mode" so that all output representations will only have the X and Y coordinates. This is useful for force OGC-compliant output (since OGC only specifies 2-D geometries).
Enhanced: 2.0.0 support for Polyhedral surfaces was introduced.
Changed: 2.1.0. Up to 2.0.x this was called ST_Force_2D.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports 3d and will not drop the z-index.
SELECT ST_AsEWKT(ST_Force2D(ST_GeomFromEWKT('CIRCULARSTRING(1 1 2, 2 3 2, 4 5 2, 6 7 2, 5 6 2)'))); st_asewkt ------------------------------------- CIRCULARSTRING(1 1,2 3,4 5,6 7,5 6) SELECT ST_AsEWKT(ST_Force2D('POLYGON((0 0 2,0 5 2,5 0 2,0 0 2),(1 1 2,3 1 2,1 3 2,1 1 2))')); st_asewkt ---------------------------------------------- POLYGON((0 0,0 5,5 0,0 0),(1 1,3 1,1 3,1 1))
ST_Force3D — Forces the geometries into XYZ mode. This is an alias for ST_Force3DZ.
geometry ST_Force3D(
geometry geomA)
;
Forces the geometries into XYZ mode. This is an alias for ST_Force_3DZ. If a geometry has no Z component, then a 0 Z coordinate is tacked on.
Enhanced: 2.0.0 support for Polyhedral surfaces was introduced.
Changed: 2.1.0. Up to 2.0.x this was called ST_Force_3D.
This function supports Polyhedral surfaces.
This method supports Circular Strings and Curves
This function supports 3d and will not drop the z-index.
--Nothing happens to an already 3D geometry SELECT ST_AsEWKT(ST_Force3D(ST_GeomFromEWKT('CIRCULARSTRING(1 1 2, 2 3 2, 4 5 2, 6 7 2, 5 6 2)'))); st_asewkt ----------------------------------------------- CIRCULARSTRING(1 1 2,2 3 2,4 5 2,6 7 2,5 6 2) SELECT ST_AsEWKT(ST_Force3D('POLYGON((0 0,0 5,5 0,0 0),(1 1,3 1,1 3,1 1))')); st_asewkt -------------------------------------------------------------- POLYGON((0 0 0,0 5 0,5 0 0,0 0 0),(1 1 0,3 1 0,1 3 0,1 1 0))
ST_Force3DZ — Forces the geometries into XYZ mode. This is a synonym for ST_Force3D.
geometry ST_Force3DZ(
geometry geomA)
;
Forces the geometries into XYZ mode. This is a synonym for ST_Force3DZ. If a geometry has no Z component, then a 0 Z coordinate is tacked on.
Enhanced: 2.0.0 support for Polyhedral surfaces was introduced.
Changed: 2.1.0. Up to 2.0.x this was called ST_Force_3DZ.
This function supports Polyhedral surfaces.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
--Nothing happens to an already 3D geometry SELECT ST_AsEWKT(ST_Force3DZ(ST_GeomFromEWKT('CIRCULARSTRING(1 1 2, 2 3 2, 4 5 2, 6 7 2, 5 6 2)'))); st_asewkt ----------------------------------------------- CIRCULARSTRING(1 1 2,2 3 2,4 5 2,6 7 2,5 6 2) SELECT ST_AsEWKT(ST_Force3DZ('POLYGON((0 0,0 5,5 0,0 0),(1 1,3 1,1 3,1 1))')); st_asewkt -------------------------------------------------------------- POLYGON((0 0 0,0 5 0,5 0 0,0 0 0),(1 1 0,3 1 0,1 3 0,1 1 0))
ST_Force3DM — Forces the geometries into XYM mode.
geometry ST_Force3DM(
geometry geomA)
;
Forces the geometries into XYM mode. If a geometry has no M component, then a 0 M coordinate is tacked on. If it has a Z component, then Z is removed
Changed: 2.1.0. Up to 2.0.x this was called ST_Force_3DM.
This method supports Circular Strings and Curves
--Nothing happens to an already 3D geometry SELECT ST_AsEWKT(ST_Force3DM(ST_GeomFromEWKT('CIRCULARSTRING(1 1 2, 2 3 2, 4 5 2, 6 7 2, 5 6 2)'))); st_asewkt ------------------------------------------------ CIRCULARSTRINGM(1 1 0,2 3 0,4 5 0,6 7 0,5 6 0) SELECT ST_AsEWKT(ST_Force3DM('POLYGON((0 0 1,0 5 1,5 0 1,0 0 1),(1 1 1,3 1 1,1 3 1,1 1 1))')); st_asewkt --------------------------------------------------------------- POLYGONM((0 0 0,0 5 0,5 0 0,0 0 0),(1 1 0,3 1 0,1 3 0,1 1 0))
ST_Force4D — Forces the geometries into XYZM mode.
geometry ST_Force4D(
geometry geomA)
;
Forces the geometries into XYZM mode. 0 is tacked on for missing Z and M dimensions.
Changed: 2.1.0. Up to 2.0.x this was called ST_Force_4D.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
--Nothing happens to an already 3D geometry SELECT ST_AsEWKT(ST_Force4D(ST_GeomFromEWKT('CIRCULARSTRING(1 1 2, 2 3 2, 4 5 2, 6 7 2, 5 6 2)'))); st_asewkt --------------------------------------------------------- CIRCULARSTRING(1 1 2 0,2 3 2 0,4 5 2 0,6 7 2 0,5 6 2 0) SELECT ST_AsEWKT(ST_Force4D('MULTILINESTRINGM((0 0 1,0 5 2,5 0 3,0 0 4),(1 1 1,3 1 1,1 3 1,1 1 1))')); st_asewkt -------------------------------------------------------------------------------------- MULTILINESTRING((0 0 0 1,0 5 0 2,5 0 0 3,0 0 0 4),(1 1 0 1,3 1 0 1,1 3 0 1,1 1 0 1))
ST_ForceCollection — Converts the geometry into a GEOMETRYCOLLECTION.
geometry ST_ForceCollection(
geometry geomA)
;
Converts the geometry into a GEOMETRYCOLLECTION. This is useful for simplifying the WKB representation.
Enhanced: 2.0.0 support for Polyhedral surfaces was introduced.
Availability: 1.2.2, prior to 1.3.4 this function will crash with Curves. This is fixed in 1.3.4+
Changed: 2.1.0. Up to 2.0.x this was called ST_Force_Collection.
This function supports Polyhedral surfaces.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_AsEWKT(ST_ForceCollection('POLYGON((0 0 1,0 5 1,5 0 1,0 0 1),(1 1 1,3 1 1,1 3 1,1 1 1))')); st_asewkt ---------------------------------------------------------------------------------- GEOMETRYCOLLECTION(POLYGON((0 0 1,0 5 1,5 0 1,0 0 1),(1 1 1,3 1 1,1 3 1,1 1 1))) SELECT ST_AsText(ST_ForceCollection('CIRCULARSTRING(220227 150406,2220227 150407,220227 150406)')); st_astext -------------------------------------------------------------------------------- GEOMETRYCOLLECTION(CIRCULARSTRING(220227 150406,2220227 150407,220227 150406)) (1 row)
-- POLYHEDRAL example -- SELECT ST_AsEWKT(ST_ForceCollection('POLYHEDRALSURFACE(((0 0 0,0 0 1,0 1 1,0 1 0,0 0 0)), ((0 0 0,0 1 0,1 1 0,1 0 0,0 0 0)), ((0 0 0,1 0 0,1 0 1,0 0 1,0 0 0)), ((1 1 0,1 1 1,1 0 1,1 0 0,1 1 0)), ((0 1 0,0 1 1,1 1 1,1 1 0,0 1 0)), ((0 0 1,1 0 1,1 1 1,0 1 1,0 0 1)))')) st_asewkt ---------------------------------------------------------------------------------- GEOMETRYCOLLECTION( POLYGON((0 0 0,0 0 1,0 1 1,0 1 0,0 0 0)), POLYGON((0 0 0,0 1 0,1 1 0,1 0 0,0 0 0)), POLYGON((0 0 0,1 0 0,1 0 1,0 0 1,0 0 0)), POLYGON((1 1 0,1 1 1,1 0 1,1 0 0,1 1 0)), POLYGON((0 1 0,0 1 1,1 1 1,1 1 0,0 1 0)), POLYGON((0 0 1,1 0 1,1 1 1,0 1 1,0 0 1)) )
ST_ForceSFS — Forces the geometries to use SFS 1.1 geometry types only.
geometry ST_ForceSFS(
geometry geomA)
;
geometry ST_ForceSFS(
geometry geomA, text version)
;
ST_ForceRHR — Forces the orientation of the vertices in a polygon to follow the Right-Hand-Rule.
boolean ST_ForceRHR(
geometry g)
;
Forces the orientation of the vertices in a polygon to follow the Right-Hand-Rule. In GIS terminology, this means that the area that is bounded by the polygon is to the right of the boundary. In particular, the exterior ring is orientated in a clockwise direction and the interior rings in a counter-clockwise direction.
Enhanced: 2.0.0 support for Polyhedral surfaces was introduced.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
ST_ForceCurve — Upcasts a geometry into its curved type, if applicable.
geometry ST_ForceCurve(
geometry g)
;
Turns a geometry into its curved representation, if applicable: lines become compoundcurves, multilines become multicurves polygons become curvepolygons multipolygons become multisurfaces. If the geometry input is already a curved representation returns back same as input.
Availability: 2.2.0
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
ST_LineMerge — Returns a (set of) LineString(s) formed by sewing together a MULTILINESTRING.
geometry ST_LineMerge(
geometry amultilinestring)
;
Returns a (set of) LineString(s) formed by sewing together the constituent line work of a MULTILINESTRING.
Only use with MULTILINESTRING/LINESTRINGs. If you feed a polygon or geometry collection into this function, it will return an empty GEOMETRYCOLLECTION |
Availability: 1.1.0
richiede GEOS >= 2.1.0 |
SELECT ST_AsText(ST_LineMerge( ST_GeomFromText('MULTILINESTRING((-29 -27,-30 -29.7,-36 -31,-45 -33),(-45 -33,-46 -32))') ) ); st_astext -------------------------------------------------------------------------------------------------- LINESTRING(-29 -27,-30 -29.7,-36 -31,-45 -33,-46 -32) (1 row) --If can't be merged - original MULTILINESTRING is returned SELECT ST_AsText(ST_LineMerge( ST_GeomFromText('MULTILINESTRING((-29 -27,-30 -29.7,-36 -31,-45 -33),(-45.2 -33.2,-46 -32))') ) ); st_astext ---------------- MULTILINESTRING((-45.2 -33.2,-46 -32),(-29 -27,-30 -29.7,-36 -31,-45 -33))
ST_CollectionExtract — Given a (multi)geometry, returns a (multi)geometry consisting only of elements of the specified type.
geometry ST_CollectionExtract(
geometry collection, integer type)
;
Given a (multi)geometry, returns a (multi)geometry consisting only of elements of the specified type. Sub-geometries that are not the specified type are ignored. If there are no sub-geometries of the right type, an EMPTY geometry will be returned. Only points, lines and polygons are supported. Type numbers are 1 == POINT, 2 == LINESTRING, 3 == POLYGON.
Availability: 1.5.0
Prior to 1.5.3 this function returned non-collection inputs untouched, no matter type. In 1.5.3 non-matching single geometries result in a NULL return. In of 2.0.0 every case of missing match results in a typed EMPTY return. |
When specifying 3 == POLYGON a multipolygon is returned even when the edges are shared. This results in an invalid multipolygon for many cases such as applying this function on an ST_Split result. |
-- constanti: 1 == POINT, 2 == LINESTRING, 3 == POLYGON SELECT ST_AsText(ST_CollectionExtract(ST_GeomFromText('GEOMETRYCOLLECTION(GEOMETRYCOLLECTION(POINT(0 0)))'),1)); st_astext --------------- MULTIPOINT(0 0) (1 row) SELECT ST_AsText(ST_CollectionExtract(ST_GeomFromText('GEOMETRYCOLLECTION(GEOMETRYCOLLECTION(LINESTRING(0 0, 1 1)),LINESTRING(2 2, 3 3))'),2)); st_astext --------------- MULTILINESTRING((0 0, 1 1), (2 2, 3 3)) (1 row)
ST_CollectionHomogenize — Given a geometry collection, returns the "simplest" representation of the contents.
geometry ST_CollectionHomogenize(
geometry collection)
;
Given a geometry collection, returns the "simplest" representation of the contents. Singletons will be returned as singletons. Collections that are homogeneous will be returned as the appropriate multi-type.
When specifying 3 == POLYGON a multipolygon is returned even when the edges are shared. This results in an invalid multipolygon for many cases such as applying this function on an ST_Split result. |
Availability: 2.0.0
ST_Multi — Returns the geometry as a MULTI* geometry. If the geometry is already a MULTI*, it is returned unchanged.
geometry ST_Multi(
geometry g1)
;
Returns the geometry as a MULTI* geometry. If the geometry is already a MULTI*, it is returned unchanged.
SELECT ST_AsText(ST_Multi(ST_GeomFromText('POLYGON((743238 2967416,743238 2967450, 743265 2967450,743265.625 2967416,743238 2967416))'))); st_astext -------------------------------------------------------------------------------------------------- MULTIPOLYGON(((743238 2967416,743238 2967450,743265 2967450,743265.625 2967416, 743238 2967416))) (1 row)
ST_RemovePoint — Removes point from a linestring. Offset is 0-based.
geometry ST_RemovePoint(
geometry linestring, integer offset)
;
Removes point from a linestring. Useful for turning a closed ring into an open line string
Availability: 1.1.0
This function supports 3d and will not drop the z-index.
ST_Reverse — Returns the geometry with vertex order reversed.
geometry ST_Reverse(
geometry g1)
;
ST_Rotate — Rotate a geometry rotRadians counter-clockwise about an origin.
geometry ST_Rotate(
geometry geomA, float rotRadians)
;
geometry ST_Rotate(
geometry geomA, float rotRadians, float x0, float y0)
;
geometry ST_Rotate(
geometry geomA, float rotRadians, geometry pointOrigin)
;
Rotates geometry rotRadians counter-clockwise about the origin. The rotation origin can be specified either as a POINT geometry, or as x and y coordinates. If the origin is not specified, the geometry is rotated about POINT(0 0).
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Enhanced: 2.0.0 additional parameters for specifying the origin of rotation were added.
Availability: 1.1.2. Name changed from Rotate to ST_Rotate in 1.2.2
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
--Rotate 180 degrees SELECT ST_AsEWKT(ST_Rotate('LINESTRING (50 160, 50 50, 100 50)', pi())); st_asewkt --------------------------------------- LINESTRING(-50 -160,-50 -50,-100 -50) (1 row) --Rotate 30 degrees counter-clockwise at x=50, y=160 SELECT ST_AsEWKT(ST_Rotate('LINESTRING (50 160, 50 50, 100 50)', pi()/6, 50, 160)); st_asewkt --------------------------------------------------------------------------- LINESTRING(50 160,105 64.7372055837117,148.301270189222 89.7372055837117) (1 row) --Rotate 60 degrees clockwise from centroid SELECT ST_AsEWKT(ST_Rotate(geom, -pi()/3, ST_Centroid(geom))) FROM (SELECT 'LINESTRING (50 160, 50 50, 100 50)'::geometry AS geom) AS foo; st_asewkt -------------------------------------------------------------- LINESTRING(116.4225 130.6721,21.1597 75.6721,46.1597 32.3708) (1 row)
ST_RotateX — Rotate a geometry rotRadians about the X axis.
geometry ST_RotateX(
geometry geomA, float rotRadians)
;
Rotate a geometry geomA - rotRadians about the X axis.
|
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Availability: 1.1.2. Name changed from RotateX to ST_RotateX in 1.2.2
This function supports Polyhedral surfaces.
This function supports 3d and will not drop the z-index.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_RotateY — Rotate a geometry rotRadians about the Y axis.
geometry ST_RotateY(
geometry geomA, float rotRadians)
;
Rotate a geometry geomA - rotRadians about the y axis.
|
Availability: 1.1.2. Name changed from RotateY to ST_RotateY in 1.2.2
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
This function supports Polyhedral surfaces.
This function supports 3d and will not drop the z-index.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_RotateZ — Rotate a geometry rotRadians about the Z axis.
geometry ST_RotateZ(
geometry geomA, float rotRadians)
;
Rotate a geometry geomA - rotRadians about the Z axis.
This is a synonym for ST_Rotate |
|
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Availability: 1.1.2. Name changed from RotateZ to ST_RotateZ in 1.2.2
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
--Rotate a line 90 degrees along z-axis SELECT ST_AsEWKT(ST_RotateZ(ST_GeomFromEWKT('LINESTRING(1 2 3, 1 1 1)'), pi()/2)); st_asewkt --------------------------- LINESTRING(-2 1 3,-1 1 1) --Rotate a curved circle around z-axis SELECT ST_AsEWKT(ST_RotateZ(the_geom, pi()/2)) FROM (SELECT ST_LineToCurve(ST_Buffer(ST_GeomFromText('POINT(234 567)'), 3)) As the_geom) As foo; st_asewkt ---------------------------------------------------------------------------------------------------------------------------- CURVEPOLYGON(CIRCULARSTRING(-567 237,-564.87867965644 236.12132034356,-564 234,-569.12132034356 231.87867965644,-567 237))
ST_Scale — Scales the geometry to a new size by multiplying the ordinates with the parameters. Ie: ST_Scale(geom, Xfactor, Yfactor, Zfactor).
geometry ST_Scale(
geometry geomA, float XFactor, float YFactor, float ZFactor)
;
geometry ST_Scale(
geometry geomA, float XFactor, float YFactor)
;
geometry ST_Scale(
geometry geom, geometry factor)
;
Scales the geometry to a new size by multiplying the ordinates with the corresponding factor parameters.
The version taking a geometry as the factor
parameter allows passing a 2d, 3dm, 3dz or 4d point to set scaling factor for all supported dimensions. Missing dimensions in the factor
point are equivalent to no scaling the corresponding dimension.
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
Availability: 1.1.0.
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Enhanced: 2.2.0 support for scaling all dimension (geometry parameter) was introduced.
This function supports Polyhedral surfaces.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This function supports M coordinates.
--Version 1: scale X, Y, Z SELECT ST_AsEWKT(ST_Scale(ST_GeomFromEWKT('LINESTRING(1 2 3, 1 1 1)'), 0.5, 0.75, 0.8)); st_asewkt -------------------------------------- LINESTRING(0.5 1.5 2.4,0.5 0.75 0.8) --Version 2: Scale X Y SELECT ST_AsEWKT(ST_Scale(ST_GeomFromEWKT('LINESTRING(1 2 3, 1 1 1)'), 0.5, 0.75)); st_asewkt ---------------------------------- LINESTRING(0.5 1.5 3,0.5 0.75 1) --Version 3: Scale X Y Z M SELECT ST_AsEWKT(ST_Scale(ST_GeomFromEWKT('LINESTRING(1 2 3 4, 1 1 1 1)'), ST_MakePoint(0.5, 0.75, 2, -1))); st_asewkt ---------------------------------------- LINESTRING(0.5 1.5 6 -4,0.5 0.75 2 -1)
ST_Segmentize — Return a modified geometry/geography having no segment longer than the given distance. Distance computation is performed in 2d only. For geometry, length units are in units of spatial reference. For geography, units are in meters.
geometry ST_Segmentize(
geometry geom, float max_segment_length)
;
geography ST_Segmentize(
geography geog, float max_segment_length)
;
Returns a modified geometry having no segment longer than the given max_segment_length
. Distance computation is performed in 2d only. For geometry, length units are in units of spatial reference. For geography, units are in meters.
Availability: 1.2.2
Enhanced: 2.1.0 support for geography was introduced.
Changed: 2.1.0 As a result of the introduction of geography support: The construct SELECT ST_Segmentize('LINESTRING(1 2, 3 4)',0.5);
will result in ambiguous function error. You need to have properly typed object e.g. a geometry/geography column, use ST_GeomFromText, ST_GeogFromText or SELECT ST_Segmentize('LINESTRING(1 2, 3 4)'::geometry,0.5);
This will only increase segments. It will not lengthen segments shorter than max length |
SELECT ST_AsText(ST_Segmentize( ST_GeomFromText('MULTILINESTRING((-29 -27,-30 -29.7,-36 -31,-45 -33),(-45 -33,-46 -32))') ,5) ); st_astext -------------------------------------------------------------------------------------------------- MULTILINESTRING((-29 -27,-30 -29.7,-34.886615700134 -30.758766735029,-36 -31, -40.8809353009198 -32.0846522890933,-45 -33), (-45 -33,-46 -32)) (1 row) SELECT ST_AsText(ST_Segmentize(ST_GeomFromText('POLYGON((-29 28, -30 40, -29 28))'),10)); st_astext ----------------------- POLYGON((-29 28,-29.8304547985374 37.9654575824488,-30 40,-29.1695452014626 30.0345424175512,-29 28)) (1 row)
ST_SetPoint — Replace point N of linestring with given point. Index is 0-based.
geometry ST_SetPoint(
geometry linestring, integer zerobasedposition, geometry point)
;
Replace point N of linestring with given point. Index is 0-based. This is especially useful in triggers when trying to maintain relationship of joints when one vertex moves.
Availability: 1.1.0
This function supports 3d and will not drop the z-index.
--Change first point in line string from -1 3 to -1 1 SELECT ST_AsText(ST_SetPoint('LINESTRING(-1 2,-1 3)', 0, 'POINT(-1 1)')); st_astext ----------------------- LINESTRING(-1 1,-1 3) ---Change last point in a line string (lets play with 3d linestring this time) SELECT ST_AsEWKT(ST_SetPoint(foo.the_geom, ST_NumPoints(foo.the_geom) - 1, ST_GeomFromEWKT('POINT(-1 1 3)'))) FROM (SELECT ST_GeomFromEWKT('LINESTRING(-1 2 3,-1 3 4, 5 6 7)') As the_geom) As foo; st_asewkt ----------------------- LINESTRING(-1 2 3,-1 3 4,-1 1 3)
ST_SetSRID — Sets the SRID on a geometry to a particular integer value.
geometry ST_SetSRID(
geometry geom, integer srid)
;
Sets the SRID on a geometry to a particular integer value. Useful in constructing bounding boxes for queries.
This function does not transform the geometry coordinates in any way - it simply sets the meta data defining the spatial reference system the geometry is assumed to be in. Use ST_Transform if you want to transform the geometry into a new projection. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method supports Circular Strings and Curves
-- Mark a point as WGS 84 long lat --
SELECT ST_SetSRID(ST_Point(-123.365556, 48.428611),4326) As wgs84long_lat; -- the ewkt representation (wrap with ST_AsEWKT) - SRID=4326;POINT(-123.365556 48.428611)
-- Mark a point as WGS 84 long lat and then transform to web mercator (Spherical Mercator) --
SELECT ST_Transform(ST_SetSRID(ST_Point(-123.365556, 48.428611),4326),3785) As spere_merc; -- the ewkt representation (wrap with ST_AsEWKT) - SRID=3785;POINT(-13732990.8753491 6178458.96425423)
ST_SnapToGrid — Snap all points of the input geometry to a regular grid.
geometry ST_SnapToGrid(
geometry geomA, float originX, float originY, float sizeX, float sizeY)
;
geometry ST_SnapToGrid(
geometry geomA, float sizeX, float sizeY)
;
geometry ST_SnapToGrid(
geometry geomA, float size)
;
geometry ST_SnapToGrid(
geometry geomA, geometry pointOrigin, float sizeX, float sizeY, float sizeZ, float sizeM)
;
Variant 1,2,3: Snap all points of the input geometry to the grid defined by its origin and cell size. Remove consecutive points falling on the same cell, eventually returning NULL if output points are not enough to define a geometry of the given type. Collapsed geometries in a collection are stripped from it. Useful for reducing precision.
Variant 4: Introduced 1.1.0 - Snap all points of the input geometry to the grid defined by its origin (the second argument, must be a point) and cell sizes. Specify 0 as size for any dimension you don't want to snap to a grid.
The returned geometry might lose its simplicity (see ST_IsSimple). |
Before release 1.1.0 this function always returned a 2d geometry. Starting at 1.1.0 the returned geometry will have same dimensionality as the input one with higher dimension values untouched. Use the version taking a second geometry argument to define all grid dimensions. |
Availability: 1.0.0RC1
Availability: 1.1.0 - Z and M support
This function supports 3d and will not drop the z-index.
--Snap your geometries to a precision grid of 10^-3 UPDATE mytable SET the_geom = ST_SnapToGrid(the_geom, 0.001); SELECT ST_AsText(ST_SnapToGrid( ST_GeomFromText('LINESTRING(1.1115678 2.123, 4.111111 3.2374897, 4.11112 3.23748667)'), 0.001) ); st_astext ------------------------------------- LINESTRING(1.112 2.123,4.111 3.237) --Snap a 4d geometry SELECT ST_AsEWKT(ST_SnapToGrid( ST_GeomFromEWKT('LINESTRING(-1.1115678 2.123 2.3456 1.11111, 4.111111 3.2374897 3.1234 1.1111, -1.11111112 2.123 2.3456 1.1111112)'), ST_GeomFromEWKT('POINT(1.12 2.22 3.2 4.4444)'), 0.1, 0.1, 0.1, 0.01) ); st_asewkt ------------------------------------------------------------------------------ LINESTRING(-1.08 2.12 2.3 1.1144,4.12 3.22 3.1 1.1144,-1.08 2.12 2.3 1.1144) --With a 4d geometry - the ST_SnapToGrid(geom,size) only touches x and y coords but keeps m and z the same SELECT ST_AsEWKT(ST_SnapToGrid(ST_GeomFromEWKT('LINESTRING(-1.1115678 2.123 3 2.3456, 4.111111 3.2374897 3.1234 1.1111)'), 0.01) ); st_asewkt --------------------------------------------------------- LINESTRING(-1.11 2.12 3 2.3456,4.11 3.24 3.1234 1.1111)
ST_Snap — Snap segments and vertices of input geometry to vertices of a reference geometry.
geometry ST_Snap(
geometry input, geometry reference, float tolerance)
;
Snaps the vertices and segments of a geometry another Geometry's vertices. A snap distance tolerance is used to control where snapping is performed.
Snapping one geometry to another can improve robustness for overlay operations by eliminating nearly-coincident edges (which cause problems during noding and intersection calculation).
Too much snapping can result in invalid topology being created, so the number and location of snapped vertices is decided using heuristics to determine when it is safe to snap. This can result in some potential snaps being omitted, however.
The returned geometry might lose its simplicity (see ST_IsSimple) and validity (see ST_IsValid). |
Availability: 2.0.0 requires GEOS >= 3.3.0.
SELECT ST_AsText(ST_Snap(poly,line, ST_Distance(poly,line)*1.01)) AS polysnapped FROM (SELECT ST_GeomFromText('MULTIPOLYGON( ((26 125, 26 200, 126 200, 126 125, 26 125 ), ( 51 150, 101 150, 76 175, 51 150 )), (( 151 100, 151 200, 176 175, 151 100 )))') As poly, ST_GeomFromText('LINESTRING (5 107, 54 84, 101 100)') As line ) As foo; polysnapped --------------------------------------------------------------------- MULTIPOLYGON(((26 125,26 200,126 200,126 125,101 100,26 125), (51 150,101 150,76 175,51 150)),((151 100,151 200,176 175,151 100))) |
SELECT ST_AsText( ST_Snap(poly,line, ST_Distance(poly,line)*1.25) ) AS polysnapped FROM (SELECT ST_GeomFromText('MULTIPOLYGON( (( 26 125, 26 200, 126 200, 126 125, 26 125 ), ( 51 150, 101 150, 76 175, 51 150 )), (( 151 100, 151 200, 176 175, 151 100 )))') As poly, ST_GeomFromText('LINESTRING (5 107, 54 84, 101 100)') As line ) As foo; polysnapped --------------------------------------------------------------------- MULTIPOLYGON(((5 107,26 200,126 200,126 125,101 100,54 84,5 107), (51 150,101 150,76 175,51 150)),((151 100,151 200,176 175,151 100))) |
SELECT ST_AsText( ST_Snap(line, poly, ST_Distance(poly,line)*1.01) ) AS linesnapped FROM (SELECT ST_GeomFromText('MULTIPOLYGON( ((26 125, 26 200, 126 200, 126 125, 26 125), (51 150, 101 150, 76 175, 51 150 )), ((151 100, 151 200, 176 175, 151 100)))') As poly, ST_GeomFromText('LINESTRING (5 107, 54 84, 101 100)') As line ) As foo; linesnapped ---------------------------------------- LINESTRING(5 107,26 125,54 84,101 100)
|
SELECT ST_AsText( ST_Snap(line, poly, ST_Distance(poly,line)*1.25) ) AS linesnapped FROM (SELECT ST_GeomFromText('MULTIPOLYGON( (( 26 125, 26 200, 126 200, 126 125, 26 125 ), (51 150, 101 150, 76 175, 51 150 )), ((151 100, 151 200, 176 175, 151 100 )))') As poly, ST_GeomFromText('LINESTRING (5 107, 54 84, 101 100)') As line ) As foo; linesnapped --------------------------------------- LINESTRING(26 125,54 84,101 100) |
ST_Transform — Returns a new geometry with its coordinates transformed to the SRID referenced by the integer parameter.
geometry ST_Transform(
geometry g1, integer srid)
;
Returns a new geometry with its coordinates transformed to spatial reference system referenced by the SRID integer parameter. The destination SRID must exist in the SPATIAL_REF_SYS
table.
ST_Transform is often confused with ST_SetSRID(). ST_Transform actually changes the coordinates of a geometry from one spatial reference system to another, while ST_SetSRID() simply changes the SRID identifier of the geometry
Requires PostGIS be compiled with Proj support. Use PostGIS_Full_Version to confirm you have proj support compiled in. |
If using more than one transformation, it is useful to have a functional index on the commonly used transformations to take advantage of index usage. |
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
Enhanced: 2.0.0 support for Polyhedral surfaces was introduced.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.6
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
Change Mass state plane US feet geometry to WGS 84 long lat
SELECT ST_AsText(ST_Transform(ST_GeomFromText('POLYGON((743238 2967416,743238 2967450, 743265 2967450,743265.625 2967416,743238 2967416))',2249),4326)) As wgs_geom; wgs_geom --------------------------- POLYGON((-71.1776848522251 42.3902896512902,-71.1776843766326 42.3903829478009, -71.1775844305465 42.3903826677917,-71.1775825927231 42.3902893647987,-71.177684 8522251 42.3902896512902)); (1 row) --3D Circular String example SELECT ST_AsEWKT(ST_Transform(ST_GeomFromEWKT('SRID=2249;CIRCULARSTRING(743238 2967416 1,743238 2967450 2,743265 2967450 3,743265.625 2967416 3,743238 2967416 4)'),4326)); st_asewkt -------------------------------------------------------------------------------------- SRID=4326;CIRCULARSTRING(-71.1776848522251 42.3902896512902 1,-71.1776843766326 42.3903829478009 2, -71.1775844305465 42.3903826677917 3, -71.1775825927231 42.3902893647987 3,-71.1776848522251 42.3902896512902 4)
Example of creating a partial functional index. For tables where you are not sure all the geometries will be filled in, its best to use a partial index that leaves out null geometries which will both conserve space and make your index smaller and more efficient.
CREATE INDEX idx_the_geom_26986_parcels ON parcels USING gist (ST_Transform(the_geom, 26986)) WHERE the_geom IS NOT NULL;
Sometimes coordinate transformation involving a grid-shift can fail, for example if PROJ.4 has not been built with grid-shift files or the coordinate does not lie within the range for which the grid shift is defined. By default, PostGIS will throw an error if a grid shift file is not present, but this behaviour can be configured on a per-SRID basis by altering the proj4text value within the spatial_ref_sys table.
For example, the proj4text parameter +datum=NAD87 is a shorthand form for the following +nadgrids parameter:
+nadgrids=@conus,@alaska,@ntv2_0.gsb,@ntv1_can.dat
The @ prefix means no error is reported if the files are not present, but if the end of the list is reached with no file having been appropriate (ie. found and overlapping) then an error is issued.
If, conversely, you wanted to ensure that at least the standard files were present, but that if all files were scanned without a hit a null transformation is applied you could use:
+nadgrids=@conus,@alaska,@ntv2_0.gsb,@ntv1_can.dat,null
The null grid shift file is a valid grid shift file covering the whole world and applying no shift. So for a complete example, if you wanted to alter PostGIS so that transformations to SRID 4267 that didn't lie within the correct range did not throw an ERROR, you would use the following:
UPDATE spatial_ref_sys SET proj4text = '+proj=longlat +ellps=clrk66 +nadgrids=@conus,@alaska,@ntv2_0.gsb,@ntv1_can.dat,null +no_defs' WHERE srid = 4267;
ST_Translate — Translates the geometry to a new location using the numeric parameters as offsets. Ie: ST_Translate(geom, X, Y) or ST_Translate(geom, X, Y,Z).
geometry ST_Translate(
geometry g1, float deltax, float deltay)
;
geometry ST_Translate(
geometry g1, float deltax, float deltay, float deltaz)
;
Returns a new geometry whose coordinates are translated delta x,delta y,delta z units. Units are based on the units defined in spatial reference (SRID) for this geometry.
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
Availability: 1.2.2
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
Move a point 1 degree longitude
SELECT ST_AsText(ST_Translate(ST_GeomFromText('POINT(-71.01 42.37)',4326),1,0)) As wgs_transgeomtxt; wgs_transgeomtxt --------------------- POINT(-70.01 42.37)
Move a linestring 1 degree longitude and 1/2 degree latitude
SELECT ST_AsText(ST_Translate(ST_GeomFromText('LINESTRING(-71.01 42.37,-71.11 42.38)',4326),1,0.5)) As wgs_transgeomtxt; wgs_transgeomtxt --------------------------------------- LINESTRING(-70.01 42.87,-70.11 42.88)
Move a 3d point
SELECT ST_AsEWKT(ST_Translate(CAST('POINT(0 0 0)' As geometry), 5, 12,3)); st_asewkt --------- POINT(5 12 3)
Move a curve and a point
SELECT ST_AsText(ST_Translate(ST_Collect('CURVEPOLYGON(CIRCULARSTRING(4 3,3.12 0.878,1 0,-1.121 5.1213,6 7, 8 9,4 3))','POINT(1 3)'),1,2)); st_astext ------------------------------------------------------------------------------------------------------------ GEOMETRYCOLLECTION(CURVEPOLYGON(CIRCULARSTRING(5 5,4.12 2.878,2 2,-0.121 7.1213,7 9,9 11,5 5)),POINT(2 5))
ST_TransScale — Translates the geometry using the deltaX and deltaY args, then scales it using the XFactor, YFactor args, working in 2D only.
geometry ST_TransScale(
geometry geomA, float deltaX, float deltaY, float XFactor, float YFactor)
;
Translates the geometry using the deltaX and deltaY args, then scales it using the XFactor, YFactor args, working in 2D only.
|
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
Availability: 1.1.0.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_AsEWKT(ST_TransScale(ST_GeomFromEWKT('LINESTRING(1 2 3, 1 1 1)'), 0.5, 1, 1, 2)); st_asewkt ----------------------------- LINESTRING(1.5 6 3,1.5 4 1) --Buffer a point to get an approximation of a circle, convert to curve and then translate 1,2 and scale it 3,4 SELECT ST_AsText(ST_Transscale(ST_LineToCurve(ST_Buffer('POINT(234 567)', 3)),1,2,3,4)); st_astext ------------------------------------------------------------------------------------------------------------------------------ CURVEPOLYGON(CIRCULARSTRING(714 2276,711.363961030679 2267.51471862576,705 2264,698.636038969321 2284.48528137424,714 2276))
ST_AsBinary — Return the Well-Known Binary (WKB) representation of the geometry/geography without SRID meta data.
bytea ST_AsBinary(
geometry g1)
;
bytea ST_AsBinary(
geometry g1, text NDR_or_XDR)
;
bytea ST_AsBinary(
geography g1)
;
bytea ST_AsBinary(
geography g1, text NDR_or_XDR)
;
Returns the Well-Known Binary representation of the geometry. There are 2 variants of the function. The first variant takes no endian encoding parameter and defaults to server machine endian. The second variant takes a second argument denoting the encoding - using little-endian ('NDR') or big-endian ('XDR') encoding.
This is useful in binary cursors to pull data out of the database without converting it to a string representation.
The WKB spec does not include the SRID. To get the WKB with SRID format use ST_AsEWKB |
ST_AsBinary is the reverse of ST_GeomFromWKB for geometry. Use ST_GeomFromWKB to convert to a postgis geometry from ST_AsBinary representation. |
The default behavior in PostgreSQL 9.0 has been changed to output bytea in hex encoding. ST_AsBinary is the reverse of ST_GeomFromWKB for geometry. If your GUI tools require the old behavior, then SET bytea_output='escape' in your database. |
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Enhanced: 2.0.0 support for higher coordinate dimensions was introduced.
Enhanced: 2.0.0 support for specifying endian with geography was introduced.
Availability: 1.5.0 geography support was introduced.
Changed: 2.0.0 Inputs to this function can not be unknown -- must be geometry. Constructs such as ST_AsBinary('POINT(1 2)')
are no longer valid and you will get an n st_asbinary(unknown) is not unique error
. Code like that needs to be changed to ST_AsBinary('POINT(1 2)'::geometry);
. If that is not possible, then install legacy.sql
.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.1
This method implements the SQL/MM specification. SQL-MM 3: 5.1.37
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This function supports 3d and will not drop the z-index.
SELECT ST_AsBinary(ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326)); st_asbinary -------------------------------- \001\003\000\000\000\001\000\000\000\005 \000\000\000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000 \000\000\000\360?\000\000\000\000\000\000 \360?\000\000\000\000\000\000\360?\000\000 \000\000\000\000\360?\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000 (1 row)
SELECT ST_AsBinary(ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326), 'XDR'); st_asbinary -------------------------------- \000\000\000\000\003\000\000\000\001\000\000\000\005\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000 \000?\360\000\000\000\000\000\000?\360\000\000\000\000\000\000?\360\000\000 \000\000\000\000?\360\000\000\000\000\000\000\000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000 (1 row)
ST_AsEWKB — Return the Well-Known Binary (WKB) representation of the geometry with SRID meta data.
bytea ST_AsEWKB(
geometry g1)
;
bytea ST_AsEWKB(
geometry g1, text NDR_or_XDR)
;
Returns the Well-Known Binary representation of the geometry with SRID metadata. There are 2 variants of the function. The first variant takes no endian encoding parameter and defaults to little endian. The second variant takes a second argument denoting the encoding - using little-endian ('NDR') or big-endian ('XDR') encoding.
This is useful in binary cursors to pull data out of the database without converting it to a string representation.
The WKB spec does not include the SRID. To get the OGC WKB format use ST_AsBinary |
ST_AsEWKB is the reverse of ST_GeomFromEWKB. Use ST_GeomFromEWKB to convert to a postgis geometry from ST_AsEWKB representation. |
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT ST_AsEWKB(ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326)); st_asewkb -------------------------------- \001\003\000\000 \346\020\000\000\001\000 \000\000\005\000\000\000\000 \000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000 \000\000\360?\000\000\000\000\000\000\360? \000\000\000\000\000\000\360?\000\000\000\000\000 \000\360?\000\000\000\000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000\000\000\000 (1 row)
SELECT ST_AsEWKB(ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326), 'XDR'); st_asewkb -------------------------------- \000 \000\000\003\000\000\020\346\000\000\000\001\000\000\000\005\000\000\000\000\ 000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000? \360\000\000\000\000\000\000?\360\000\000\000\000\000\000?\360\000\000\000\000 \000\000?\360\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000 \000\000\000\000\000\000\000\000\000\000\000\000\000
ST_AsEWKT — Return the Well-Known Text (WKT) representation of the geometry with SRID meta data.
text ST_AsEWKT(
geometry g1)
;
text ST_AsEWKT(
geography g1)
;
Returns the Well-Known Text representation of the geometry prefixed with the SRID.
The WKT spec does not include the SRID. To get the OGC WKT format use ST_AsText |
WKT format does not maintain precision so to prevent floating truncation, use ST_AsBinary or ST_AsEWKB format for transport.
ST_AsEWKT is the reverse of ST_GeomFromEWKT. Use ST_GeomFromEWKT to convert to a postgis geometry from ST_AsEWKT representation. |
Enhanced: 2.0.0 support for Geography, Polyhedral surfaces, Triangles and TIN was introduced.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT ST_AsEWKT('0103000020E61000000100000005000000000000 000000000000000000000000000000000000000000000000000000 F03F000000000000F03F000000000000F03F000000000000F03 F000000000000000000000000000000000000000000000000'::geometry); st_asewkt -------------------------------- SRID=4326;POLYGON((0 0,0 1,1 1,1 0,0 0)) (1 row) SELECT ST_AsEWKT('0108000080030000000000000060E30A4100000000785C0241000000000000F03F0000000018 E20A4100000000485F024100000000000000400000000018 E20A4100000000305C02410000000000000840') --st_asewkt--- CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)
ST_AsGeoJSON — Restisuisce la geometria come elemento GeoJSON.
text ST_AsGeoJSON(
geometry geom, integer maxdecimaldigits=15, integer options=0)
;
text ST_AsGeoJSON(
geography geog, integer maxdecimaldigits=15, integer options=0)
;
text ST_AsGeoJSON(
integer gj_version, geometry geom, integer maxdecimaldigits=15, integer options=0)
;
text ST_AsGeoJSON(
integer gj_version, geography geog, integer maxdecimaldigits=15, integer options=0)
;
Return the geometry as a Geometry Javascript Object Notation (GeoJSON) element. (Cf GeoJSON specifications 1.0). 2D and 3D Geometries are both supported. GeoJSON only support SFS 1.1 geometry type (no curve support for example).
The gj_version parameter is the major version of the GeoJSON spec. If specified, must be 1. This represents the spec version of GeoJSON.
The third argument may be used to reduce the maximum number of decimal places used in output (defaults to 15).
The last 'options' argument could be used to add Bbox or Crs in GeoJSON output:
0: means no option (default value)
1: GeoJSON Bbox
2: GeoJSON Short CRS (e.g EPSG:4326)
4: GeoJSON Long CRS (e.g urn:ogc:def:crs:EPSG::4326)
Version 1: ST_AsGeoJSON(geom) / precision=15 version=1 options=0
Version 2: ST_AsGeoJSON(geom, precision) / version=1 options=0
Version 3: ST_AsGeoJSON(geom, precision, options) / version=1
Version 4: ST_AsGeoJSON(gj_version, geom) / precision=15 options=0
Version 5: ST_AsGeoJSON(gj_version, geom, precision) /options=0
Version 6: ST_AsGeoJSON(gj_version, geom, precision,options)
Disponibilità: 1.3.4
Availability: 1.5.0 geography support was introduced.
Changed: 2.0.0 support default args and named args.
This function supports 3d and will not drop the z-index.
GeoJSON format is generally more efficient than other formats for use in ajax mapping. One popular javascript client that supports this is Open Layers. Example of its use is OpenLayers GeoJSON Example
SELECT ST_AsGeoJSON(the_geom) from fe_edges limit 1; st_asgeojson ----------------------------------------------------------------------------------------------------------- {"type":"MultiLineString","coordinates":[[[-89.734634999999997,31.492072000000000], [-89.734955999999997,31.492237999999997]]]} (1 row) --3d point SELECT ST_AsGeoJSON('LINESTRING(1 2 3, 4 5 6)'); st_asgeojson ----------------------------------------------------------------------------------------- {"type":"LineString","coordinates":[[1,2,3],[4,5,6]]}
ST_AsGML — Return the geometry as a GML version 2 or 3 element.
text ST_AsGML(
geometry geom, integer maxdecimaldigits=15, integer options=0)
;
text ST_AsGML(
geography geog, integer maxdecimaldigits=15, integer options=0)
;
text ST_AsGML(
integer version, geometry geom, integer maxdecimaldigits=15, integer options=0, text nprefix=null, text id=null)
;
text ST_AsGML(
integer version, geography geog, integer maxdecimaldigits=15, integer options=0, text nprefix=null, text id=null)
;
Return the geometry as a Geography Markup Language (GML) element. The version parameter, if specified, may be either 2 or 3. If no version parameter is specified then the default is assumed to be 2. The precision argument may be used to reduce the maximum number of decimal places (maxdecimaldigits
) used in output (defaults to 15).
GML 2 refer to 2.1.2 version, GML 3 to 3.1.1 version
The 'options' argument is a bitfield. It could be used to define CRS output type in GML output, and to declare data as lat/lon:
0: GML Short CRS (e.g EPSG:4326), default value
1: GML Long CRS (e.g urn:ogc:def:crs:EPSG::4326)
2: For GML 3 only, remove srsDimension attribute from output.
4: For GML 3 only, use <LineString> rather than <Curve> tag for lines.
16: Declare that datas are lat/lon (e.g srid=4326). Default is to assume that data are planars. This option is useful for GML 3.1.1 output only, related to axis order. So if you set it, it will swap the coordinates so order is lat lon instead of database lon lat.
32: Output the box of the geometry (envelope).
The 'namespace prefix' argument may be used to specify a custom namespace prefix or no prefix (if empty). If null or omitted 'gml' prefix is used
Disponibilità: 1.3.2
Availability: 1.5.0 geography support was introduced.
Enhanced: 2.0.0 prefix support was introduced. Option 4 for GML3 was introduced to allow using LineString instead of Curve tag for lines. GML3 Support for Polyhedral surfaces and TINS was introduced. Option 32 was introduced to output the box.
Changed: 2.0.0 use default named args
Enhanced: 2.1.0 id support was introduced, for GML 3.
Only version 3+ of ST_AsGML supports Polyhedral Surfaces and TINS. |
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT ST_AsGML(ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326)); st_asgml -------- <gml:Polygon srsName="EPSG:4326" ><gml:outerBoundaryIs ><gml:LinearRing ><gml:coordinates >0,0 0,1 1,1 1,0 0,0</gml:coordinates ></gml:LinearRing ></gml:outerBoundaryIs ></gml:Polygon >
-- Flip coordinates and output extended EPSG (16 | 1)-- SELECT ST_AsGML(3, ST_GeomFromText('POINT(5.234234233242 6.34534534534)',4326), 5, 17); st_asgml -------- <gml:Point srsName="urn:ogc:def:crs:EPSG::4326"><gml:pos>6.34535 5.23423</gml:pos></gml:Point>
-- Output the envelope (32) -- SELECT ST_AsGML(3, ST_GeomFromText('LINESTRING(1 2, 3 4, 10 20)',4326), 5, 32); st_asgml -------- <gml:Envelope srsName="EPSG:4326"> <gml:lowerCorner>1 2</gml:lowerCorner> <gml:upperCorner>10 20</gml:upperCorner> </gml:Envelope>
-- Output the envelope (32) , reverse (lat lon instead of lon lat) (16), long srs (1)= 32 | 16 | 1 = 49 -- SELECT ST_AsGML(3, ST_GeomFromText('LINESTRING(1 2, 3 4, 10 20)',4326), 5, 49); st_asgml -------- <gml:Envelope srsName="urn:ogc:def:crs:EPSG::4326"> <gml:lowerCorner>2 1</gml:lowerCorner> <gml:upperCorner>20 10</gml:upperCorner> </gml:Envelope>
-- Polyhedral Example -- SELECT ST_AsGML(3, ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')); st_asgml -------- <gml:PolyhedralSurface> <gml:polygonPatches> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing> <gml:posList srsDimension="3">0 0 0 0 0 1 0 1 1 0 1 0 0 0 0</gml:posList> </gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing> <gml:posList srsDimension="3">0 0 0 0 1 0 1 1 0 1 0 0 0 0 0</gml:posList> </gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing> <gml:posList srsDimension="3">0 0 0 1 0 0 1 0 1 0 0 1 0 0 0</gml:posList> </gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing> <gml:posList srsDimension="3">1 1 0 1 1 1 1 0 1 1 0 0 1 1 0</gml:posList> </gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing> <gml:posList srsDimension="3">0 1 0 0 1 1 1 1 1 1 1 0 0 1 0</gml:posList> </gml:LinearRing> </gml:exterior> </gml:PolygonPatch> <gml:PolygonPatch> <gml:exterior> <gml:LinearRing> <gml:posList srsDimension="3">0 0 1 1 0 1 1 1 1 0 1 1 0 0 1</gml:posList> </gml:LinearRing> </gml:exterior> </gml:PolygonPatch> </gml:polygonPatches> </gml:PolyhedralSurface>
ST_AsHEXEWKB — Returns a Geometry in HEXEWKB format (as text) using either little-endian (NDR) or big-endian (XDR) encoding.
text ST_AsHEXEWKB(
geometry g1, text NDRorXDR)
;
text ST_AsHEXEWKB(
geometry g1)
;
Returns a Geometry in HEXEWKB format (as text) using either little-endian (NDR) or big-endian (XDR) encoding. If no encoding is specified, then NDR is used.
Availability: 1.2.2 |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_AsHEXEWKB(ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326)); dà la stessa risposta di SELECT ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326)::text; st_ashexewkb -------- 0103000020E6100000010000000500 00000000000000000000000000000000 00000000000000000000000000000000F03F 000000000000F03F000000000000F03F000000000000F03 F000000000000000000000000000000000000000000000000
ST_AsKML — Return the geometry as a KML element. Several variants. Default version=2, default precision=15
text ST_AsKML(
geometry geom, integer maxdecimaldigits=15)
;
text ST_AsKML(
geography geog, integer maxdecimaldigits=15)
;
text ST_AsKML(
integer version, geometry geom, integer maxdecimaldigits=15, text nprefix=NULL)
;
text ST_AsKML(
integer version, geography geog, integer maxdecimaldigits=15, text nprefix=NULL)
;
Return the geometry as a Keyhole Markup Language (KML) element. There are several variants of this function. maximum number of decimal places used in output (defaults to 15), version default to 2 and default namespace is no prefix.
Version 1: ST_AsKML(geom_or_geog, maxdecimaldigits) / version=2 / maxdecimaldigits=15
Version 2: ST_AsKML(version, geom_or_geog, maxdecimaldigits, nprefix) maxdecimaldigits=15 / nprefix=NULL
Requires PostGIS be compiled with Proj support. Use PostGIS_Full_Version to confirm you have proj support compiled in. |
Availability: 1.2.2 - later variants that include version param came in 1.3.2 |
Enhanced: 2.0.0 - Add prefix namespace. Default is no prefix |
Changed: 2.0.0 - uses default args and supports named args |
AsKML output will not work with geometries that do not have an SRID |
This function supports 3d and will not drop the z-index.
SELECT ST_AsKML(ST_GeomFromText('POLYGON((0 0,0 1,1 1,1 0,0 0))',4326)); st_askml -------- <Polygon ><outerBoundaryIs ><LinearRing ><coordinates >0,0 0,1 1,1 1,0 0,0</coordinates ></LinearRing ></outerBoundaryIs ></Polygon> --3d linestring SELECT ST_AsKML('SRID=4326;LINESTRING(1 2 3, 4 5 6)'); <LineString ><coordinates >1,2,3 4,5,6</coordinates ></LineString>
ST_AsSVG — Returns a Geometry in SVG path data given a geometry or geography object.
text ST_AsSVG(
geometry geom, integer rel=0, integer maxdecimaldigits=15)
;
text ST_AsSVG(
geography geog, integer rel=0, integer maxdecimaldigits=15)
;
Return the geometry as Scalar Vector Graphics (SVG) path data. Use 1 as second argument to have the path data implemented in terms of relative moves, the default (or 0) uses absolute moves. Third argument may be used to reduce the maximum number of decimal digits used in output (defaults to 15). Point geometries will be rendered as cx/cy when 'rel' arg is 0, x/y when 'rel' is 1. Multipoint geometries are delimited by commas (","), GeometryCollection geometries are delimited by semicolons (";").
Availability: 1.2.2. Availability: 1.4.0 Changed in PostGIS 1.4.0 to include L command in absolute path to conform to http://www.w3.org/TR/SVG/paths.html#PathDataBNF |
Changed: 2.0.0 to use default args and support named args
ST_AsX3D — Returns a Geometry in X3D xml node element format: ISO-IEC-19776-1.2-X3DEncodings-XML
text ST_AsX3D(
geometry g1, integer maxdecimaldigits=15, integer options=0)
;
Returns a geometry as an X3D xml formatted node element http://www.web3d.org/standards/number/19776-1. If maxdecimaldigits
(precision) is not specified then defaults to 15.
There are various options for translating PostGIS geometries to X3D since X3D geometry types don't map directly to PostGIS geometry types and some newer X3D types that might be better mappings we have avoided since most rendering tools don't currently support them. These are the mappings we have settled on. Feel free to post a bug ticket if you have thoughts on the idea or ways we can allow people to denote their preferred mappings. Below is how we currently map PostGIS 2D/3D types to X3D types |
The 'options' argument is a bitfield. For PostGIS 2.2+, this is used to denote whether to represent coordinates with X3D GeoCoordinates Geospatial node and also whether to flip the x/y axis. By default, ST_AsX3D
outputs in database form (long,lat or X,Y), but X3D default of lat/lon, y/x may be preferred.
0: X/Y in database order (e.g. long/lat = X,Y is standard database order), default value, and non-spatial coordinates (just regular old Coordinate tag).
1: Flip X and Y. If used in conjunction with the GeoCoordinate option switch, then output will be default "latitude_first" and coordinates will be flipped as well.
2: Output coordinates in GeoSpatial GeoCoordinates. This option will throw an error if geometries are not in WGS 84 long lat (srid: 4326). This is currently the only GeoCoordinate type supported. Refer to X3D specs specifying a spatial reference system.. Default output will be GeoCoordinate geoSystem='"GD" "WE" "longitude_first"'
. If you prefer the X3D default of GeoCoordinate geoSystem='"GD" "WE" "latitude_first"'
use (2 + 1)
= 3
PostGIS Type | 2D X3D Type | 3D X3D Type |
---|---|---|
LINESTRING | not yet implemented - will be PolyLine2D | LineSet |
MULTILINESTRING | not yet implemented - will be PolyLine2D | IndexedLineSet |
MULTIPOINT | Polypoint2D | PointSet |
POINT | outputs the space delimited coordinates | outputs the space delimited coordinates |
(MULTI) POLYGON, POLYHEDRALSURFACE | Invalid X3D markup | IndexedFaceSet (inner rings currently output as another faceset) |
TIN | TriangleSet2D (Not Yet Implemented) | IndexedTriangleSet |
2D geometry support not yet complete. Inner rings currently just drawn as separate polygons. We are working on these. |
Lots of advancements happening in 3D space particularly with X3D Integration with HTML5
There is also a nice open source X3D viewer you can use to view rendered geometries. Free Wrl http://freewrl.sourceforge.net/ binaries available for Mac, Linux, and Windows. Use the FreeWRL_Launcher packaged to view the geometries.
Also check out PostGIS minimalist X3D viewer that utilizes this function and x3dDom html/js open source toolkit.
Availability: 2.0.0: ISO-IEC-19776-1.2-X3DEncodings-XML
Enhanced: 2.2.0: Support for GeoCoordinates and axis (x/y, long/lat) flipping. Look at options for details.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT '<?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE X3D PUBLIC "ISO//Web3D//DTD X3D 3.0//EN" "http://www.web3d.org/specifications/x3d-3.0.dtd"> <X3D> <Scene> <Transform> <Shape> <Appearance> <Material emissiveColor=''0 0 1''/> </Appearance > ' || ST_AsX3D( ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')) || '</Shape> </Transform> </Scene> </X3D >' As x3ddoc; x3ddoc -------- <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE X3D PUBLIC "ISO//Web3D//DTD X3D 3.0//EN" "http://www.web3d.org/specifications/x3d-3.0.dtd"> <X3D> <Scene> <Transform> <Shape> <Appearance> <Material emissiveColor='0 0 1'/> </Appearance > <IndexedFaceSet coordIndex='0 1 2 3 -1 4 5 6 7 -1 8 9 10 11 -1 12 13 14 15 -1 16 17 18 19 -1 20 21 22 23'> <Coordinate point='0 0 0 0 0 1 0 1 1 0 1 0 0 0 0 0 1 0 1 1 0 1 0 0 0 0 0 1 0 0 1 0 1 0 0 1 1 1 0 1 1 1 1 0 1 1 0 0 0 1 0 0 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 0 1 1' /> </IndexedFaceSet> </Shape> </Transform> </Scene> </X3D >
SELECT ST_AsX3D( ST_Translate( ST_Force_3d( ST_Buffer(ST_Point(10,10),5, 'quad_segs=2')), 0,0, 3) ,6) As x3dfrag; x3dfrag -------- <IndexedFaceSet coordIndex="0 1 2 3 4 5 6 7"> <Coordinate point="15 10 3 13.535534 6.464466 3 10 5 3 6.464466 6.464466 3 5 10 3 6.464466 13.535534 3 10 15 3 13.535534 13.535534 3 " /> </IndexedFaceSet >
SELECT ST_AsX3D(ST_GeomFromEWKT('TIN ((( 0 0 0, 0 0 1, 0 1 0, 0 0 0 )), (( 0 0 0, 0 1 0, 1 1 0, 0 0 0 )) )')) As x3dfrag; x3dfrag -------- <IndexedTriangleSet index='0 1 2 3 4 5' ><Coordinate point='0 0 0 0 0 1 0 1 0 0 0 0 0 1 0 1 1 0'/></IndexedTriangleSet >
SELECT ST_AsX3D( ST_GeomFromEWKT('MULTILINESTRING((20 0 10,16 -12 10,0 -16 10,-12 -12 10,-20 0 10,-12 16 10,0 24 10,16 16 10,20 0 10), (12 0 10,8 8 10,0 12 10,-8 8 10,-8 0 10,-8 -4 10,0 -8 10,8 -4 10,12 0 10))') ) As x3dfrag; x3dfrag -------- <IndexedLineSet coordIndex='0 1 2 3 4 5 6 7 0 -1 8 9 10 11 12 13 14 15 8'> <Coordinate point='20 0 10 16 -12 10 0 -16 10 -12 -12 10 -20 0 10 -12 16 10 0 24 10 16 16 10 12 0 10 8 8 10 0 12 10 -8 8 10 -8 0 10 -8 -4 10 0 -8 10 8 -4 10 ' /> </IndexedLineSet >
ST_GeoHash — Return a GeoHash representation of the geometry.
text ST_GeoHash(
geometry geom, integer maxchars=full_precision_of_point)
;
Return a GeoHash representation (http://en.wikipedia.org/wiki/Geohash) of the geometry. A GeoHash encodes a point into a text form that is sortable and searchable based on prefixing. A shorter GeoHash is a less precise representation of a point. It can also be thought of as a box, that contains the actual point.
If no maxchars
is specified ST_GeoHash returns a GeoHash based on full precision of the input geometry type. Points return a GeoHash with 20 characters of precision (about enough to hold the full double precision of the input). Other types return a GeoHash with a variable amount of precision, based on the size of the feature. Larger features are represented with less precision, smaller features with more precision. The idea is that the box implied by the GeoHash will always contain the input feature.
If maxchars
is specified ST_GeoHash returns a GeoHash with at most that many characters so a possibly lower precision representation of the input geometry. For non-points, the starting point of the calculation is the center of the bounding box of the geometry.
Disponibilità: 1.4.0
ST_GeoHash will not work with geometries that are not in geographic (lon/lat) coordinates. |
This method supports Circular Strings and Curves
ST_AsText — Return the Well-Known Text (WKT) representation of the geometry/geography without SRID metadata.
text ST_AsText(
geometry g1)
;
text ST_AsText(
geography g1)
;
Returns the Well-Known Text representation of the geometry/geography.
The WKT spec does not include the SRID. To get the SRID as part of the data, use the non-standard PostGIS ST_AsEWKT |
WKT format does not maintain precision so to prevent floating truncation, use ST_AsBinary or ST_AsEWKB format for transport.
ST_AsText is the reverse of ST_GeomFromText. Use ST_GeomFromText to convert to a postgis geometry from ST_AsText representation. |
Availability: 1.5 - support for geography was introduced.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.1
This method implements the SQL/MM specification. SQL-MM 3: 5.1.25
This method supports Circular Strings and Curves
ST_AsLatLonText — Return the Degrees, Minutes, Seconds representation of the given point.
text ST_AsLatLonText(
geometry pt, text format='')
;
Returns the Degrees, Minutes, Seconds representation of the point.
It is assumed the point is in a lat/lon projection. The X (lon) and Y (lat) coordinates are normalized in the output to the "normal" range (-180 to +180 for lon, -90 to +90 for lat). |
The text parameter is a format string containing the format for the resulting text, similar to a date format string. Valid tokens are "D" for degrees, "M" for minutes, "S" for seconds, and "C" for cardinal direction (NSEW). DMS tokens may be repeated to indicate desired width and precision ("SSS.SSSS" means " 1.0023").
"M", "S", and "C" are optional. If "C" is omitted, degrees are shown with a "-" sign if south or west. If "S" is omitted, minutes will be shown as decimal with as many digits of precision as you specify. If "M" is also omitted, degrees are shown as decimal with as many digits precision as you specify.
If the format string is omitted (or zero-length) a default format will be used.
Availability: 2.0
Default format.
SELECT (ST_AsLatLonText('POINT (-3.2342342 -2.32498)')); st_aslatlontext ---------------------------- 2°19'29.928"S 3°14'3.243"W
Providing a format (same as the default).
SELECT (ST_AsLatLonText('POINT (-3.2342342 -2.32498)', 'D°M''S.SSS"C')); st_aslatlontext ---------------------------- 2°19'29.928"S 3°14'3.243"W
Characters other than D, M, S, C and . are just passed through.
SELECT (ST_AsLatLonText('POINT (-3.2342342 -2.32498)', 'D degrees, M minutes, S seconds to the C')); st_aslatlontext -------------------------------------------------------------------------------------- 2 degrees, 19 minutes, 30 seconds to the S 3 degrees, 14 minutes, 3 seconds to the W
Signed degrees instead of cardinal directions.
SELECT (ST_AsLatLonText('POINT (-3.2342342 -2.32498)', 'D°M''S.SSS"')); st_aslatlontext ---------------------------- -2°19'29.928" -3°14'3.243"
Decimal degrees.
SELECT (ST_AsLatLonText('POINT (-3.2342342 -2.32498)', 'D.DDDD degrees C')); st_aslatlontext ----------------------------------- 2.3250 degrees S 3.2342 degrees W
Excessively large values are normalized.
SELECT (ST_AsLatLonText('POINT (-302.2342342 -792.32498)')); st_aslatlontext ------------------------------- 72°19'29.928"S 57°45'56.757"E
ST_AsTWKB — Returns the geometry as TWKB, aka "Tiny Well-Known Binary"
bytea ST_AsTWKB(
geometry g1, integer decimaldigits_xy=0, integer decimaldigits_z=0, integer decimaldigits_m=0, boolean include_sizes=false, boolean include_bounding boxes=false)
;
bytea ST_AsTWKB(
geometry[] geometries, bigint[] unique_ids, integer decimaldigits_xy=0, integer decimaldigits_z=0, integer decimaldigits_m=0, boolean include_sizes=false, boolean include_bounding_boxes=false)
;
Returns the geometry in TWKB (Tiny Well-Known Binary) format. TWKB is a compressed binary format with a focus on minimizing the size of the output.
The decimal digits parameters control how much precision is stored in the output. By default, values are rounded to the nearest unit before encoding. If you want to transfer more precision, increase the number. For example, a value of 1 implies that the first digit to the right of the decimal point will be preserved.
The sizes and bounding boxes parameters control whether optional information about the encoded length of the object and the bounds of the object are included in the output. By default they are not. Do not turn them on unless your client software has a use for them, as they just use up space (and saving space is the point of TWKB).
The array-input form of the function is used to convert a collection of geometries and unique identifiers into a TWKB collection that preserves the identifiers. This is useful for clients that expect to unpack a collection and then access further information about the objects inside. You can create the arrays using the array_agg function. The other parameters operate the same as for the simple form of the function.
The format specification is available online at https://github.com/TWKB/Specification, and code for building a JavaScript client can be found at https://github.com/TWKB/twkb.js. |
Availability: 2.2.0
SELECT ST_AsTWKB('LINESTRING(1 1,5 5)'::geometry); st_astwkb -------------------------------------------- \x02000202020808
To create an aggregate TWKB object including identifiers aggregate the desired geometries and objects first, using "array_agg()", then call the appropriate TWKB function.
SELECT ST_AsTWKB(array_agg(geom), array_agg(gid)) FROM mytable; st_astwkb -------------------------------------------- \x040402020400000202
TRUE
if A's 2D bounding box intersects B's 2D bounding box.TRUE
if A's n-D bounding box intersects B's n-D bounding box.TRUE
if A's bounding box overlaps or is to the left of B's.TRUE
if A's bounding box overlaps or is below B's.TRUE
if A' bounding box overlaps or is to the right of B's.TRUE
if A's bounding box is strictly to the left of B's.TRUE
if A's bounding box is strictly below B's.TRUE
if A's bounding box is the same as B's. Uses double precision bounding box.TRUE
if A's bounding box is strictly to the right of B's.TRUE
if A's bounding box is contained by B's.TRUE
if A's bounding box overlaps or is above B's.TRUE
if A's bounding box is strictly above B's.TRUE
if A's bounding box contains B's.TRUE
if A's bounding box is the same as B's.&& — Returns TRUE
if A's 2D bounding box intersects B's 2D bounding box.
boolean &&(
geometry A , geometry B )
;
boolean &&(
geography A , geography B )
;
The &&
operator returns TRUE
if the 2D bounding box of geometry A intersects the 2D bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
Enhanced: 2.0.0 support for Polyhedral surfaces was introduced.
Availability: 1.5.0 support for geography was introduced.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
SELECT tbl1.column1, tbl2.column1, tbl1.column2 && tbl2.column2 AS overlaps FROM ( VALUES (1, 'LINESTRING(0 0, 3 3)'::geometry), (2, 'LINESTRING(0 1, 0 5)'::geometry)) AS tbl1, ( VALUES (3, 'LINESTRING(1 2, 4 6)'::geometry)) AS tbl2; column1 | column1 | overlaps ---------+---------+---------- 1 | 3 | t 2 | 3 | f (2 rows)
&&& — Returns TRUE
if A's n-D bounding box intersects B's n-D bounding box.
boolean &&&(
geometry A , geometry B )
;
The &&&
operator returns TRUE
if the n-D bounding box of geometry A intersects the n-D bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
Disponibilità: 2.0
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This function supports 3d and will not drop the z-index.
SELECT tbl1.column1, tbl2.column1, tbl1.column2 &&& tbl2.column2 AS overlaps_3d, tbl1.column2 && tbl2.column2 AS overlaps_2d FROM ( VALUES (1, 'LINESTRING Z(0 0 1, 3 3 2)'::geometry), (2, 'LINESTRING Z(1 2 0, 0 5 -1)'::geometry)) AS tbl1, ( VALUES (3, 'LINESTRING Z(1 2 1, 4 6 1)'::geometry)) AS tbl2; column1 | column1 | overlaps_3d | overlaps_2d ---------+---------+-------------+------------- 1 | 3 | t | t 2 | 3 | f | t
SELECT tbl1.column1, tbl2.column1, tbl1.column2 &&& tbl2.column2 AS overlaps_3zm, tbl1.column2 && tbl2.column2 AS overlaps_2d FROM ( VALUES (1, 'LINESTRING M(0 0 1, 3 3 2)'::geometry), (2, 'LINESTRING M(1 2 0, 0 5 -1)'::geometry)) AS tbl1, ( VALUES (3, 'LINESTRING M(1 2 1, 4 6 1)'::geometry)) AS tbl2; column1 | column1 | overlaps_3zm | overlaps_2d ---------+---------+-------------+------------- 1 | 3 | t | t 2 | 3 | f | t
&< — Returns TRUE
if A's bounding box overlaps or is to the left of B's.
boolean &<(
geometry A , geometry B )
;
The &<
operator returns TRUE
if the bounding box of geometry A overlaps or is to the left of the bounding box of geometry B, or more accurately, overlaps or is NOT to the right of the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 &< tbl2.column2 AS overleft FROM ( VALUES (1, 'LINESTRING(1 2, 4 6)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING(0 0, 3 3)'::geometry), (3, 'LINESTRING(0 1, 0 5)'::geometry), (4, 'LINESTRING(6 0, 6 1)'::geometry)) AS tbl2; column1 | column1 | overleft ---------+---------+---------- 1 | 2 | f 1 | 3 | f 1 | 4 | t (3 rows)
&<| — Returns TRUE
if A's bounding box overlaps or is below B's.
boolean &<|(
geometry A , geometry B )
;
The &<|
operator returns TRUE
if the bounding box of geometry A overlaps or is below of the bounding box of geometry B, or more accurately, overlaps or is NOT above the bounding box of geometry B.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 &<| tbl2.column2 AS overbelow FROM ( VALUES (1, 'LINESTRING(6 0, 6 4)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING(0 0, 3 3)'::geometry), (3, 'LINESTRING(0 1, 0 5)'::geometry), (4, 'LINESTRING(1 2, 4 6)'::geometry)) AS tbl2; column1 | column1 | overbelow ---------+---------+----------- 1 | 2 | f 1 | 3 | t 1 | 4 | t (3 rows)
&> — Returns TRUE
if A' bounding box overlaps or is to the right of B's.
boolean &>(
geometry A , geometry B )
;
The &>
operator returns TRUE
if the bounding box of geometry A overlaps or is to the right of the bounding box of geometry B, or more accurately, overlaps or is NOT to the left of the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 &> tbl2.column2 AS overright FROM ( VALUES (1, 'LINESTRING(1 2, 4 6)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING(0 0, 3 3)'::geometry), (3, 'LINESTRING(0 1, 0 5)'::geometry), (4, 'LINESTRING(6 0, 6 1)'::geometry)) AS tbl2; column1 | column1 | overright ---------+---------+----------- 1 | 2 | t 1 | 3 | t 1 | 4 | f (3 rows)
<< — Returns TRUE
if A's bounding box is strictly to the left of B's.
boolean <<(
geometry A , geometry B )
;
The <<
operator returns TRUE
if the bounding box of geometry A is strictly to the left of the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 << tbl2.column2 AS left FROM ( VALUES (1, 'LINESTRING (1 2, 1 5)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING (0 0, 4 3)'::geometry), (3, 'LINESTRING (6 0, 6 5)'::geometry), (4, 'LINESTRING (2 2, 5 6)'::geometry)) AS tbl2; column1 | column1 | left ---------+---------+------ 1 | 2 | f 1 | 3 | t 1 | 4 | t (3 rows)
<<| — Returns TRUE
if A's bounding box is strictly below B's.
boolean <<|(
geometry A , geometry B )
;
The <<|
operator returns TRUE
if the bounding box of geometry A is strictly below the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 <<| tbl2.column2 AS below FROM ( VALUES (1, 'LINESTRING (0 0, 4 3)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING (1 4, 1 7)'::geometry), (3, 'LINESTRING (6 1, 6 5)'::geometry), (4, 'LINESTRING (2 3, 5 6)'::geometry)) AS tbl2; column1 | column1 | below ---------+---------+------- 1 | 2 | t 1 | 3 | f 1 | 4 | f (3 rows)
= — Returns TRUE
if A's bounding box is the same as B's. Uses double precision bounding box.
boolean =(
geometry A , geometry B )
;
boolean =(
geography A , geography B )
;
The =
operator returns TRUE
if the bounding box of geometry/geography A is the same as the bounding box of geometry/geography B. PostgreSQL uses the =, <, and > operators defined for geometries to perform internal orderings and comparison of geometries (ie. in a GROUP BY or ORDER BY clause).
This is cause for a lot of confusion. When you compare geometryA = geometryB it will return true even when the geometries are clearly different IF their bounding boxes are the same. To check for true equality use ST_OrderingEquals or ST_Equals |
This operand will NOT make use of any indexes that may be available on the geometries. |
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
Changed: 2.0.0 , the bounding box of geometries was changed to use double precision instead of float4 precision of prior. The side effect of this is that in particular points in prior versions that were a little different may have returned true in prior versions and false in 2.0+ since their float4 boxes would be the same but there float8 (double precision), would be different.
SELECT 'LINESTRING(0 0, 0 1, 1 0)'::geometry = 'LINESTRING(1 1, 0 0)'::geometry; ?column? ---------- t (1 row) SELECT ST_AsText(column1) FROM ( VALUES ('LINESTRING(0 0, 1 1)'::geometry), ('LINESTRING(1 1, 0 0)'::geometry)) AS foo; st_astext --------------------- LINESTRING(0 0,1 1) LINESTRING(1 1,0 0) (2 rows) -- Nota: il GROUP BY usa lo "=" per confrontare l'equivalenza delle geometrie SELECT ST_AsText(column1) FROM ( VALUES ('LINESTRING(0 0, 1 1)'::geometry), ('LINESTRING(1 1, 0 0)'::geometry)) AS foo GROUP BY column1; st_astext --------------------- LINESTRING(0 0,1 1) (1 row) -- In versions prior to 2.0, this used to return true -- SELECT ST_GeomFromText('POINT(1707296.37 4820536.77)') = ST_GeomFromText('POINT(1707296.27 4820536.87)') As pt_intersect; --pt_intersect -- f
>> — Returns TRUE
if A's bounding box is strictly to the right of B's.
boolean >>(
geometry A , geometry B )
;
The >>
operator returns TRUE
if the bounding box of geometry A is strictly to the right of the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 >> tbl2.column2 AS right FROM ( VALUES (1, 'LINESTRING (2 3, 5 6)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING (1 4, 1 7)'::geometry), (3, 'LINESTRING (6 1, 6 5)'::geometry), (4, 'LINESTRING (0 0, 4 3)'::geometry)) AS tbl2; column1 | column1 | right ---------+---------+------- 1 | 2 | t 1 | 3 | f 1 | 4 | f (3 rows)
@ — Returns TRUE
if A's bounding box is contained by B's.
boolean @(
geometry A , geometry B )
;
The @
operator returns TRUE
if the bounding box of geometry A is completely contained by the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 @ tbl2.column2 AS contained FROM ( VALUES (1, 'LINESTRING (1 1, 3 3)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING (0 0, 4 4)'::geometry), (3, 'LINESTRING (2 2, 4 4)'::geometry), (4, 'LINESTRING (1 1, 3 3)'::geometry)) AS tbl2; column1 | column1 | contained ---------+---------+----------- 1 | 2 | t 1 | 3 | f 1 | 4 | t (3 rows)
|&> — Returns TRUE
if A's bounding box overlaps or is above B's.
boolean |&>(
geometry A , geometry B )
;
The |&>
operator returns TRUE
if the bounding box of geometry A overlaps or is above the bounding box of geometry B, or more accurately, overlaps or is NOT below the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 |&> tbl2.column2 AS overabove FROM ( VALUES (1, 'LINESTRING(6 0, 6 4)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING(0 0, 3 3)'::geometry), (3, 'LINESTRING(0 1, 0 5)'::geometry), (4, 'LINESTRING(1 2, 4 6)'::geometry)) AS tbl2; column1 | column1 | overabove ---------+---------+----------- 1 | 2 | t 1 | 3 | f 1 | 4 | f (3 rows)
|>> — Returns TRUE
if A's bounding box is strictly above B's.
boolean |>>(
geometry A , geometry B )
;
The |>>
operator returns TRUE
if the bounding box of geometry A is strictly to the right of the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 |>> tbl2.column2 AS above FROM ( VALUES (1, 'LINESTRING (1 4, 1 7)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING (0 0, 4 2)'::geometry), (3, 'LINESTRING (6 1, 6 5)'::geometry), (4, 'LINESTRING (2 3, 5 6)'::geometry)) AS tbl2; column1 | column1 | above ---------+---------+------- 1 | 2 | t 1 | 3 | f 1 | 4 | f (3 rows)
~ — Returns TRUE
if A's bounding box contains B's.
boolean ~(
geometry A , geometry B )
;
The ~
operator returns TRUE
if the bounding box of geometry A completely contains the bounding box of geometry B.
This operand will make use of any indexes that may be available on the geometries. |
SELECT tbl1.column1, tbl2.column1, tbl1.column2 ~ tbl2.column2 AS contains FROM ( VALUES (1, 'LINESTRING (0 0, 3 3)'::geometry)) AS tbl1, ( VALUES (2, 'LINESTRING (0 0, 4 4)'::geometry), (3, 'LINESTRING (1 1, 2 2)'::geometry), (4, 'LINESTRING (0 0, 3 3)'::geometry)) AS tbl2; column1 | column1 | contains ---------+---------+---------- 1 | 2 | f 1 | 3 | t 1 | 4 | t (3 rows)
~= — Returns TRUE
if A's bounding box is the same as B's.
boolean ~=(
geometry A , geometry B )
;
The ~=
operator returns TRUE
if the bounding box of geometry/geography A is the same as the bounding box of geometry/geography B.
This operand will make use of any indexes that may be available on the geometries. |
Availability: 1.5.0 changed behavior
This function supports Polyhedral surfaces.
This operator has changed behavior in PostGIS 1.5 from testing for actual geometric equality to only checking for bounding box equality. To complicate things it also depends on if you have done a hard or soft upgrade which behavior your database has. To find out which behavior your database has you can run the query below. To check for true equality use ST_OrderingEquals or ST_Equals and to check for bounding box equality =; operator is a safer option. |
<-> — Returns the 2D distance between A and B.
double precision <->(
geometry A , geometry B )
;
double precision <->(
geography A , geography B )
;
The <->
operator returns the 2D distance between two geometries. Used in the "ORDER BY" clause provides index-assisted nearest-neighbor result sets. For PostgreSQL below 9.5 only gives centroid distance of bounding boxes and for PostgreSQL 9.5+, does true KNN distance search giving true distance between geometries, and distance sphere for geographies.
This operand will make use of 2D GiST indexes that may be available on the geometries. It is different from other operators that use spatial indexes in that the spatial index is only used when the operator is in the ORDER BY clause. |
Index only kicks in if one of the geometries is a constant (not in a subquery/cte). e.g. 'SRID=3005;POINT(1011102 450541)'::geometry instead of a.geom |
Refer to OpenGeo workshop: Nearest-Neighbour Searching for real live example.
Enhanced: 2.2.0 -- True KNN ("K nearest neighbor") behavior for geometry and geography for PostgreSQL 9.5+. Note for geography KNN is based on sphere rather than spheroid. For PostgreSQL 9.4 and below, geography support is new but only supports centroid box.
Changed: 2.2.0 -- For PostgreSQL 9.5 users, old Hybrid syntax may be slower, so you'll want to get rid of that hack if you are running your code only on PostGIS 2.2+ 9.5+. See examples below.
Availability: 2.0.0 -- Weak KNN provides nearest neighbors based on geometry centroid distances instead of true distances. Exact results for points, inexact for all other types. Available for PostgreSQL 9.1+
SELECT ST_Distance(geom, 'SRID=3005;POINT(1011102 450541)'::geometry) as d,edabbr, vaabbr FROM va2005 ORDER BY d limit 10; d | edabbr | vaabbr ------------------+--------+-------- 0 | ALQ | 128 5541.57712511724 | ALQ | 129A 5579.67450712005 | ALQ | 001 6083.4207708641 | ALQ | 131 7691.2205404848 | ALQ | 003 7900.75451037313 | ALQ | 122 8694.20710669982 | ALQ | 129B 9564.24289057111 | ALQ | 130 12089.665931705 | ALQ | 127 18472.5531479404 | ALQ | 002 (10 rows)
Then the KNN raw answer:
SELECT st_distance(geom, 'SRID=3005;POINT(1011102 450541)'::geometry) as d,edabbr, vaabbr FROM va2005 ORDER BY geom <-> 'SRID=3005;POINT(1011102 450541)'::geometry limit 10; d | edabbr | vaabbr ------------------+--------+-------- 0 | ALQ | 128 5541.57712511724 | ALQ | 129A 5579.67450712005 | ALQ | 001 6083.4207708641 | ALQ | 131 7691.2205404848 | ALQ | 003 7900.75451037313 | ALQ | 122 8694.20710669982 | ALQ | 129B 9564.24289057111 | ALQ | 130 12089.665931705 | ALQ | 127 18472.5531479404 | ALQ | 002 (10 rows)
If you run "EXPLAIN ANALYZE" on the two queries you would see a performance improvement for the second.
For users running with PostgreSQL < 9.5, use a hybrid query to find the true nearest neighbors. First a CTE query using the index-assisted KNN, then an exact query to get correct ordering:
WITH index_query AS ( SELECT ST_Distance(geom, 'SRID=3005;POINT(1011102 450541)'::geometry) as d,edabbr, vaabbr FROM va2005 ORDER BY geom <-> 'SRID=3005;POINT(1011102 450541)'::geometry LIMIT 100) SELECT * FROM index_query ORDER BY d limit 10; d | edabbr | vaabbr ------------------+--------+-------- 0 | ALQ | 128 5541.57712511724 | ALQ | 129A 5579.67450712005 | ALQ | 001 6083.4207708641 | ALQ | 131 7691.2205404848 | ALQ | 003 7900.75451037313 | ALQ | 122 8694.20710669982 | ALQ | 129B 9564.24289057111 | ALQ | 130 12089.665931705 | ALQ | 127 18472.5531479404 | ALQ | 002 (10 rows)
|=| — Returns the distance between A and B trajectories at their closest point of approach.
double precision |=|(
geometry A , geometry B )
;
The |=|
operator returns the 3D distance between two trajectories (See ST_IsValidTrajectory). This is the same as ST_DistanceCPA but as an operator it can be used for doing nearest neightbor searches using an N-dimensional index (requires PostgreSQL 9.5.0 or higher).
This operand will make use of ND GiST indexes that may be available on the geometries. It is different from other operators that use spatial indexes in that the spatial index is only used when the operator is in the ORDER BY clause. |
Index only kicks in if one of the geometries is a constant (not in a subquery/cte). e.g. 'SRID=3005;LINESTRINGM(0 0 0,0 0 1)'::geometry instead of a.geom |
Availability: 2.2.0. Index-supported only available for PostgreSQL 9.5+
-- Save a literal query trajectory in a psql variable... \set qt 'ST_AddMeasure(ST_MakeLine(ST_MakePointM(-350,300,0),ST_MakePointM(-410,490,0)),10,20)' -- Run the query ! SELECT track_id, dist FROM ( SELECT track_id, ST_DistanceCPA(tr,:qt) dist FROM trajectories ORDER BY tr |=| :qt LIMIT 5 ) foo; track_id dist ----------+------------------- 395 | 0.576496831518066 380 | 5.06797130410151 390 | 7.72262293958322 385 | 9.8004461358071 405 | 10.9534397988433 (5 rows)
<#> — Returns the 2D distance between A and B bounding boxes.
double precision <#>(
geometry A , geometry B )
;
The <#>
operator returns distance between two floating point bounding boxes, possibly reading them from a spatial index (PostgreSQL 9.1+ required). Useful for doing nearest neighbor approximate distance ordering.
This operand will make use of any indexes that may be available on the geometries. It is different from other operators that use spatial indexes in that the spatial index is only used when the operator is in the ORDER BY clause. |
Index only kicks in if one of the geometries is a constant e.g. ORDER BY (ST_GeomFromText('POINT(1 2)') <#> geom) instead of g1.geom <#>. |
Availability: 2.0.0 -- KNN only available for PostgreSQL 9.1+
SELECT * FROM ( SELECT b.tlid, b.mtfcc, b.geom <# > ST_GeomFromText('LINESTRING(746149 2948672,745954 2948576, 745787 2948499,745740 2948468,745712 2948438, 745690 2948384,745677 2948319)',2249) As b_dist, ST_Distance(b.geom, ST_GeomFromText('LINESTRING(746149 2948672,745954 2948576, 745787 2948499,745740 2948468,745712 2948438, 745690 2948384,745677 2948319)',2249)) As act_dist FROM bos_roads As b ORDER BY b_dist, b.tlid LIMIT 100) As foo ORDER BY act_dist, tlid LIMIT 10; tlid | mtfcc | b_dist | act_dist -----------+-------+------------------+------------------ 85732027 | S1400 | 0 | 0 85732029 | S1400 | 0 | 0 85732031 | S1400 | 0 | 0 85734335 | S1400 | 0 | 0 85736037 | S1400 | 0 | 0 624683742 | S1400 | 0 | 128.528874268666 85719343 | S1400 | 260.839270432962 | 260.839270432962 85741826 | S1400 | 164.759294123275 | 260.839270432962 85732032 | S1400 | 277.75 | 311.830282365264 85735592 | S1400 | 222.25 | 311.830282365264 (10 rows)
<<->> — Returns the n-D distance between the centroids of A and B bounding boxes.
double precision <<->>(
geometry A , geometry B )
;
The <<->>
operator returns the n-D (euclidean) distance between the centroids of the bounding boxes of two geometries. Useful for doing nearest neighbor approximate distance ordering.
This operand will make use of n-D GiST indexes that may be available on the geometries. It is different from other operators that use spatial indexes in that the spatial index is only used when the operator is in the ORDER BY clause. |
Index only kicks in if one of the geometries is a constant (not in a subquery/cte). e.g. 'SRID=3005;POINT(1011102 450541)'::geometry instead of a.geom |
Availability: 2.2.0 -- KNN only available for PostgreSQL 9.1+
<<#>> — Returns the n-D distance between A and B bounding boxes.
double precision <<#>>(
geometry A , geometry B )
;
The <<#>>
operator returns distance between two floating point bounding boxes, possibly reading them from a spatial index (PostgreSQL 9.1+ required). Useful for doing nearest neighbor approximate distance ordering.
This operand will make use of any indexes that may be available on the geometries. It is different from other operators that use spatial indexes in that the spatial index is only used when the operator is in the ORDER BY clause. |
Index only kicks in if one of the geometries is a constant e.g. ORDER BY (ST_GeomFromText('POINT(1 2)') <<#>> geom) instead of g1.geom <<#>>. |
Availability: 2.2.0 -- KNN only available for PostgreSQL 9.1+
TRUE
if the supplied geometries have some, but not all, interior points in common.ST_Length
POINT
guaranteed to lie on the surface.POINT
projected from a start point using a distance in meters and bearing (azimuth) in radians.TRUE
if the geometries have at least one point in common, but their interiors do not intersect.ST_3DClosestPoint — Returns the 3-dimensional point on g1 that is closest to g2. This is the first point of the 3D shortest line.
geometry ST_3DClosestPoint(
geometry g1, geometry g2)
;
Returns the 3-dimensional point on g1 that is closest to g2. This is the first point of the 3D shortest line. The 3D length of the 3D shortest line is the 3D distance.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
Disponibilità: 2.0
Changed: 2.2.0 - if 2 2D geometries are input, a 2D point is returned (instead of old behavior assuming 0 for missing Z). In case of 2D and 3D, Z is no longer assumed to be 0 for missing Z.
linestring and point -- both 3d and 2d closest point SELECT ST_AsEWKT(ST_3DClosestPoint(line,pt)) AS cp3d_line_pt, ST_AsEWKT(ST_ClosestPoint(line,pt)) As cp2d_line_pt FROM (SELECT 'POINT(100 100 30)'::geometry As pt, 'LINESTRING (20 80 20, 98 190 1, 110 180 3, 50 75 1000)'::geometry As line ) As foo; cp3d_line_pt | cp2d_line_pt -----------------------------------------------------------+------------------------------------------ POINT(54.6993798867619 128.935022917228 11.5475869506606) | POINT(73.0769230769231 115.384615384615)
|
linestring and multipoint -- both 3d and 2d closest point SELECT ST_AsEWKT(ST_3DClosestPoint(line,pt)) AS cp3d_line_pt, ST_AsEWKT(ST_ClosestPoint(line,pt)) As cp2d_line_pt FROM (SELECT 'MULTIPOINT(100 100 30, 50 74 1000)'::geometry As pt, 'LINESTRING (20 80 20, 98 190 1, 110 180 3, 50 75 900)'::geometry As line ) As foo; cp3d_line_pt | cp2d_line_pt -----------------------------------------------------------+-------------- POINT(54.6993798867619 128.935022917228 11.5475869506606) | POINT(50 75)
|
Multilinestring and polygon both 3d and 2d closest point SELECT ST_AsEWKT(ST_3DClosestPoint(poly, mline)) As cp3d, ST_AsEWKT(ST_ClosestPoint(poly, mline)) As cp2d FROM (SELECT ST_GeomFromEWKT('POLYGON((175 150 5, 20 40 5, 35 45 5, 50 60 5, 100 100 5, 175 150 5))') As poly, ST_GeomFromEWKT('MULTILINESTRING((175 155 2, 20 40 20, 50 60 -2, 125 100 1, 175 155 1), (1 10 2, 5 20 1))') As mline ) As foo; cp3d | cp2d -------------------------------------------+-------------- POINT(39.993580415989 54.1889925532825 5) | POINT(20 40)
|
ST_3DDistance — For geometry type Returns the 3-dimensional cartesian minimum distance (based on spatial ref) between two geometries in projected units.
float ST_3DDistance(
geometry g1, geometry g2)
;
For geometry type returns the 3-dimensional minimum cartesian distance between two geometries in projected units (spatial ref units).
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This method implements the SQL/MM specification. SQL-MM ?
This method is also provided by SFCGAL backend.
Disponibilità: 2.0
Changed: 2.2.0 - In case of 2D and 3D, Z is no longer assumed to be 0 for missing Z.
-- Geometry example - units in meters (SRID: 2163 US National Atlas Equal area) (3D point and line compared 2D point and line) -- Note: currently no vertical datum support so Z is not transformed and assumed to be same units as final. SELECT ST_3DDistance( ST_Transform(ST_GeomFromEWKT('SRID=4326;POINT(-72.1235 42.3521 4)'),2163), ST_Transform(ST_GeomFromEWKT('SRID=4326;LINESTRING(-72.1260 42.45 15, -72.123 42.1546 20)'),2163) ) As dist_3d, ST_Distance( ST_Transform(ST_GeomFromText('POINT(-72.1235 42.3521)',4326),2163), ST_Transform(ST_GeomFromText('LINESTRING(-72.1260 42.45, -72.123 42.1546)', 4326),2163) ) As dist_2d; dist_3d | dist_2d ------------------+----------------- 127.295059324629 | 126.66425605671
-- Multilinestring and polygon both 3d and 2d distance -- Same example as 3D closest point example SELECT ST_3DDistance(poly, mline) As dist3d, ST_Distance(poly, mline) As dist2d FROM (SELECT ST_GeomFromEWKT('POLYGON((175 150 5, 20 40 5, 35 45 5, 50 60 5, 100 100 5, 175 150 5))') As poly, ST_GeomFromEWKT('MULTILINESTRING((175 155 2, 20 40 20, 50 60 -2, 125 100 1, 175 155 1), (1 10 2, 5 20 1))') As mline ) As foo; dist3d | dist2d -------------------+-------- 0.716635696066337 | 0
ST_Distance, ST_3DClosestPoint, ST_3DDWithin, ST_3DMaxDistance, ST_3DShortestLine, ST_Transform
ST_3DDWithin — For 3d (z) geometry type Returns true if two geometries 3d distance is within number of units.
boolean ST_3DDWithin(
geometry g1, geometry g2, double precision distance_of_srid)
;
For geometry type returns true if the 3d distance between two objects is within distance_of_srid specified projected units (spatial ref units).
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This method implements the SQL/MM specification. SQL-MM ?
Disponibilità: 2.0
-- Geometry example - units in meters (SRID: 2163 US National Atlas Equal area) (3D point and line compared 2D point and line) -- Note: currently no vertical datum support so Z is not transformed and assumed to be same units as final. SELECT ST_3DDWithin( ST_Transform(ST_GeomFromEWKT('SRID=4326;POINT(-72.1235 42.3521 4)'),2163), ST_Transform(ST_GeomFromEWKT('SRID=4326;LINESTRING(-72.1260 42.45 15, -72.123 42.1546 20)'),2163), 126.8 ) As within_dist_3d, ST_DWithin( ST_Transform(ST_GeomFromEWKT('SRID=4326;POINT(-72.1235 42.3521 4)'),2163), ST_Transform(ST_GeomFromEWKT('SRID=4326;LINESTRING(-72.1260 42.45 15, -72.123 42.1546 20)'),2163), 126.8 ) As within_dist_2d; within_dist_3d | within_dist_2d ----------------+---------------- f | t
ST_3DDFullyWithin — Returns true if all of the 3D geometries are within the specified distance of one another.
boolean ST_3DDFullyWithin(
geometry g1, geometry g2, double precision distance)
;
Returns true if the 3D geometries are fully within the specified distance of one another. The distance is specified in units defined by the spatial reference system of the geometries. For this function to make sense, the source geometries must both be of the same coordinate projection, having the same SRID.
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. |
Disponibilità: 2.0
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
-- This compares the difference between fully within and distance within as well -- as the distance fully within for the 2D footprint of the line/point vs. the 3d fully within SELECT ST_3DDFullyWithin(geom_a, geom_b, 10) as D3DFullyWithin10, ST_3DDWithin(geom_a, geom_b, 10) as D3DWithin10, ST_DFullyWithin(geom_a, geom_b, 20) as D2DFullyWithin20, ST_3DDFullyWithin(geom_a, geom_b, 20) as D3DFullyWithin20 from (select ST_GeomFromEWKT('POINT(1 1 2)') as geom_a, ST_GeomFromEWKT('LINESTRING(1 5 2, 2 7 20, 1 9 100, 14 12 3)') as geom_b) t1; d3dfullywithin10 | d3dwithin10 | d2dfullywithin20 | d3dfullywithin20 ------------------+-------------+------------------+------------------ f | t | t | f
ST_3DIntersects — Returns TRUE if the Geometries "spatially intersect" in 3d - only for points, linestrings, polygons, polyhedral surface (area). With SFCGAL backend enabled also supports TINS
boolean ST_3DIntersects(
geometry geomA , geometry geomB )
;
Overlaps, Touches, Within all imply spatial intersection. If any of the aforementioned returns true, then the geometries also spatially intersect. Disjoint implies false for spatial intersection.
Disponibilità: 2.0
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. |
In order to take advantage of support for TINS, you need to enable the SFCGAL backend. This can be done at session time with: |
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This method is also provided by SFCGAL backend.
This method implements the SQL/MM specification. SQL-MM 3: ?
SELECT ST_3DIntersects(pt, line), ST_Intersects(pt,line) FROM (SELECT 'POINT(0 0 2)'::geometry As pt, 'LINESTRING (0 0 1, 0 2 3 )'::geometry As line) As foo; st_3dintersects | st_intersects -----------------+--------------- f | t (1 row)
ST_3DLongestLine — Returns the 3-dimensional longest line between two geometries
geometry ST_3DLongestLine(
geometry g1, geometry g2)
;
Returns the 3-dimensional longest line between two geometries. The function will only return the first longest line if more than one. The line returned will always start in g1 and end in g2. The 3D length of the line this function returns will always be the same as ST_3DMaxDistance returns for g1 and g2.
Disponibilità: 2.0
Changed: 2.2.0 - if 2 2D geometries are input, a 2D point is returned (instead of old behavior assuming 0 for missing Z). In case of 2D and 3D, Z is no longer assumed to be 0 for missing Z.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
linestring and point -- both 3d and 2d longest line SELECT ST_AsEWKT(ST_3DLongestLine(line,pt)) AS lol3d_line_pt, ST_AsEWKT(ST_LongestLine(line,pt)) As lol2d_line_pt FROM (SELECT 'POINT(100 100 30)'::geometry As pt, 'LINESTRING (20 80 20, 98 190 1, 110 180 3, 50 75 1000)'::geometry As line ) As foo; lol3d_line_pt | lol2d_line_pt -----------------------------------+---------------------------- LINESTRING(50 75 1000,100 100 30) | LINESTRING(98 190,100 100)
|
linestring and multipoint -- both 3d and 2d longest line SELECT ST_AsEWKT(ST_3DLongestLine(line,pt)) AS lol3d_line_pt, ST_AsEWKT(ST_LongestLine(line,pt)) As lol2d_line_pt FROM (SELECT 'MULTIPOINT(100 100 30, 50 74 1000)'::geometry As pt, 'LINESTRING (20 80 20, 98 190 1, 110 180 3, 50 75 900)'::geometry As line ) As foo; lol3d_line_pt | lol2d_line_pt ---------------------------------+-------------------------- LINESTRING(98 190 1,50 74 1000) | LINESTRING(98 190,50 74)
|
Multilinestring and polygon both 3d and 2d longest line SELECT ST_AsEWKT(ST_3DLongestLine(poly, mline)) As lol3d, ST_AsEWKT(ST_LongestLine(poly, mline)) As lol2d FROM (SELECT ST_GeomFromEWKT('POLYGON((175 150 5, 20 40 5, 35 45 5, 50 60 5, 100 100 5, 175 150 5))') As poly, ST_GeomFromEWKT('MULTILINESTRING((175 155 2, 20 40 20, 50 60 -2, 125 100 1, 175 155 1), (1 10 2, 5 20 1))') As mline ) As foo; lol3d | lol2d ------------------------------+-------------------------- LINESTRING(175 150 5,1 10 2) | LINESTRING(175 150,1 10)
|
ST_3DMaxDistance — For geometry type Returns the 3-dimensional cartesian maximum distance (based on spatial ref) between two geometries in projected units.
float ST_3DMaxDistance(
geometry g1, geometry g2)
;
For geometry type returns the 3-dimensional maximum cartesian distance between two geometries in projected units (spatial ref units).
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
Disponibilità: 2.0
Changed: 2.2.0 - In case of 2D and 3D, Z is no longer assumed to be 0 for missing Z.
-- Geometry example - units in meters (SRID: 2163 US National Atlas Equal area) (3D point and line compared 2D point and line) -- Note: currently no vertical datum support so Z is not transformed and assumed to be same units as final. SELECT ST_3DMaxDistance( ST_Transform(ST_GeomFromEWKT('SRID=4326;POINT(-72.1235 42.3521 10000)'),2163), ST_Transform(ST_GeomFromEWKT('SRID=4326;LINESTRING(-72.1260 42.45 15, -72.123 42.1546 20)'),2163) ) As dist_3d, ST_MaxDistance( ST_Transform(ST_GeomFromEWKT('SRID=4326;POINT(-72.1235 42.3521 10000)'),2163), ST_Transform(ST_GeomFromEWKT('SRID=4326;LINESTRING(-72.1260 42.45 15, -72.123 42.1546 20)'),2163) ) As dist_2d; dist_3d | dist_2d ------------------+------------------ 24383.7467488441 | 22247.8472107251
ST_3DShortestLine — Returns the 3-dimensional shortest line between two geometries
geometry ST_3DShortestLine(
geometry g1, geometry g2)
;
Returns the 3-dimensional shortest line between two geometries. The function will only return the first shortest line if more than one, that the function finds. If g1 and g2 intersects in just one point the function will return a line with both start and end in that intersection-point. If g1 and g2 are intersecting with more than one point the function will return a line with start and end in the same point but it can be any of the intersecting points. The line returned will always start in g1 and end in g2. The 3D length of the line this function returns will always be the same as ST_3DDistance returns for g1 and g2.
Disponibilità: 2.0
Changed: 2.2.0 - if 2 2D geometries are input, a 2D point is returned (instead of old behavior assuming 0 for missing Z). In case of 2D and 3D, Z is no longer assumed to be 0 for missing Z.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
linestring and point -- both 3d and 2d shortest line SELECT ST_AsEWKT(ST_3DShortestLine(line,pt)) AS shl3d_line_pt, ST_AsEWKT(ST_ShortestLine(line,pt)) As shl2d_line_pt FROM (SELECT 'POINT(100 100 30)'::geometry As pt, 'LINESTRING (20 80 20, 98 190 1, 110 180 3, 50 75 1000)'::geometry As line ) As foo; shl3d_line_pt | shl2d_line_pt ----------------------------------------------------------------------------+------------------------------------------------------ LINESTRING(54.6993798867619 128.935022917228 11.5475869506606,100 100 30) | LINESTRING(73.0769230769231 115.384615384615,100 100)
|
linestring and multipoint -- both 3d and 2d shortest line SELECT ST_AsEWKT(ST_3DShortestLine(line,pt)) AS shl3d_line_pt, ST_AsEWKT(ST_ShortestLine(line,pt)) As shl2d_line_pt FROM (SELECT 'MULTIPOINT(100 100 30, 50 74 1000)'::geometry As pt, 'LINESTRING (20 80 20, 98 190 1, 110 180 3, 50 75 900)'::geometry As line ) As foo; shl3d_line_pt | shl2d_line_pt ---------------------------------------------------------------------------+------------------------ LINESTRING(54.6993798867619 128.935022917228 11.5475869506606,100 100 30) | LINESTRING(50 75,50 74)
|
Multilinestring and polygon both 3d and 2d shortest line SELECT ST_AsEWKT(ST_3DShortestLine(poly, mline)) As shl3d, ST_AsEWKT(ST_ShortestLine(poly, mline)) As shl2d FROM (SELECT ST_GeomFromEWKT('POLYGON((175 150 5, 20 40 5, 35 45 5, 50 60 5, 100 100 5, 175 150 5))') As poly, ST_GeomFromEWKT('MULTILINESTRING((175 155 2, 20 40 20, 50 60 -2, 125 100 1, 175 155 1), (1 10 2, 5 20 1))') As mline ) As foo; shl3d | shl2d ---------------------------------------------------------------------------------------------------+------------------------ LINESTRING(39.993580415989 54.1889925532825 5,40.4078575708294 53.6052383805529 5.03423778139177) | LINESTRING(20 40,20 40)
|
ST_Area — Returns the area of the surface if it is a Polygon or MultiPolygon. For geometry, a 2D Cartesian area is determined with units specified by the SRID. For geography, area is determined on a curved surface with units in square meters.
float ST_Area(
geometry g1)
;
float ST_Area(
geography geog, boolean use_spheroid=true)
;
Returns the area of the geometry if it is a Polygon or MultiPolygon. Return the area measurement of an ST_Surface or ST_MultiSurface value. For geometry, a 2D Cartesian area is determined with units specified by the SRID. For geography, by default area is determined on a spheroid with units in square meters. To measure around the faster but less accurate sphere, use ST_Area(geog,false).
Enhanced: 2.0.0 - support for 2D polyhedral surfaces was introduced.
Enhanced: 2.2.0 - measurement on spheroid performed with GeographicLib for improved accuracy and robustness. Requires Proj >= 4.9.0 to take advantage of the new feature.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 8.1.2, 9.5.3
This function supports Polyhedral surfaces.
For polyhedral surfaces, only supports 2D polyhedral surfaces (not 2.5D). For 2.5D, may give a non-zero answer, but only for the faces that sit completely in XY plane. |
This method is also provided by SFCGAL backend.
Return area in square feet for a plot of Massachusetts land and multiply by conversion to get square meters. Note this is in square feet because EPSG:2249 is Massachusetts State Plane Feet
SELECT ST_Area(the_geom) As sqft, ST_Area(the_geom)*POWER(0.3048,2) As sqm FROM (SELECT ST_GeomFromText('POLYGON((743238 2967416,743238 2967450, 743265 2967450,743265.625 2967416,743238 2967416))',2249) ) As foo(the_geom); sqft | sqm ---------+------------- 928.625 | 86.27208552
Return area square feet and transform to Massachusetts state plane meters (EPSG:26986) to get square meters. Note this is in square feet because 2249 is Massachusetts State Plane Feet and transformed area is in square meters since EPSG:26986 is state plane Massachusetts meters
SELECT ST_Area(the_geom) As sqft, ST_Area(ST_Transform(the_geom,26986)) As sqm FROM (SELECT ST_GeomFromText('POLYGON((743238 2967416,743238 2967450, 743265 2967450,743265.625 2967416,743238 2967416))',2249) ) As foo(the_geom); sqft | sqm ---------+------------------ 928.625 | 86.2724304199219
Return area square feet and square meters using geography data type. Note that we transform to our geometry to geography (before you can do that make sure your geometry is in WGS 84 long lat 4326). Geography always measures in meters. This is just for demonstration to compare. Normally your table will be stored in geography data type already.
SELECT ST_Area(the_geog)/POWER(0.3048,2) As sqft_spheroid, ST_Area(the_geog,false)/POWER(0.3048,2) As sqft_sphere, ST_Area(the_geog) As sqm_spheroid FROM (SELECT geography( ST_Transform( ST_GeomFromText('POLYGON((743238 2967416,743238 2967450,743265 2967450,743265.625 2967416,743238 2967416))', 2249 ) ,4326 ) ) ) As foo(the_geog); sqft_spheroid | sqft_sphere | sqm_spheroid ------------------+------------------+------------------ 928.684403538925 | 927.049336105925 | 86.2776042893529 --if your data is in geography already SELECT ST_Area(the_geog)/POWER(0.3048,2) As sqft, ST_Area(the_geog) As sqm FROM somegeogtable;
ST_Azimuth — Returns the north-based azimuth as the angle in radians measured clockwise from the vertical on pointA to pointB.
float ST_Azimuth(
geometry pointA, geometry pointB)
;
float ST_Azimuth(
geography pointA, geography pointB)
;
Returns the azimuth in radians of the segment defined by the given point geometries, or NULL if the two points are coincident. The azimuth is angle is referenced from north, and is positive clockwise: North = 0; East = π/2; South = π; West = 3π/2.
For the geography type, the forward azimuth is solved as part of the inverse geodesic problem.
The azimuth is mathematical concept defined as the angle between a reference plane and a point, with angular units in radians. Units can be converted to degrees using a built-in PostgreSQL function degrees(), as shown in the example.
Disponibilità: 1.1.0
Enhanced: 2.0.0 support for geography was introduced.
Enhanced: 2.2.0 measurement on spheroid performed with GeographicLib for improved accuracy and robustness. Requires Proj >= 4.9.0 to take advantage of the new feature.
Azimuth is especially useful in conjunction with ST_Translate for shifting an object along its perpendicular axis. See upgis_lineshift Plpgsqlfunctions PostGIS wiki section for example of this.
ST_Centroid — Returns the geometric center of a geometry.
geometry ST_Centroid(
geometry g1)
;
Computes the geometric center of a geometry, or equivalently, the center of mass of the geometry as a POINT
. For [MULTI
]POINT
s, this is computed as the arithmetic mean of the input coordinates. For [MULTI
]LINESTRING
s, this is computed as the weighted length of each line segment. For [MULTI
]POLYGON
s, "weight" is thought in terms of area. If an empty geometry is supplied, an empty GEOMETRYCOLLECTION
is returned. If NULL
is supplied, NULL
is returned.
The centroid is equal to the centroid of the set of component Geometries of highest dimension (since the lower-dimension geometries contribute zero "weight" to the centroid).
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 8.1.4, 9.5.5
In each of the following illustrations, the blue dot represents the centroid of the source geometry.
SELECT ST_AsText(ST_Centroid('MULTIPOINT ( -1 0, -1 2, -1 3, -1 4, -1 7, 0 1, 0 3, 1 1, 2 0, 6 0, 7 8, 9 8, 10 6 )')); st_astext ------------------------------------------ POINT(2.30769230769231 3.30769230769231) (1 row)
ST_ClosestPoint — Returns the 2-dimensional point on g1 that is closest to g2. This is the first point of the shortest line.
geometry ST_ClosestPoint(
geometry g1, geometry g2)
;
Returns the 2-dimensional point on g1 that is closest to g2. This is the first point of the shortest line.
If you have a 3D Geometry, you may prefer to use ST_3DClosestPoint. |
Disponibilità: 1.5.0
SELECT ST_AsText(ST_ClosestPoint(pt,line)) AS cp_pt_line, ST_AsText(ST_ClosestPoint(line,pt)) As cp_line_pt FROM (SELECT 'POINT(100 100)'::geometry As pt, 'LINESTRING (20 80, 98 190, 110 180, 50 75 )'::geometry As line ) As foo; cp_pt_line | cp_line_pt ----------------+------------------------------------------ POINT(100 100) | POINT(73.0769230769231 115.384615384615)
|
SELECT ST_AsText( ST_ClosestPoint( ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) ) ) As ptwkt; ptwkt ------------------------------------------ POINT(140.752120669087 125.695053378061)
|
ST_Contains — Returns true if and only if no points of B lie in the exterior of A, and at least one point of the interior of B lies in the interior of A.
boolean ST_Contains(
geometry geomA, geometry geomB)
;
Geometry A contains Geometry B if and only if no points of B lie in the exterior of A, and at least one point of the interior of B lies in the interior of A. An important subtlety of this definition is that A does not contain its boundary, but A does contain itself. Contrast that to ST_ContainsProperly where geometry A does not Contain Properly itself.
Returns TRUE if geometry B is completely inside geometry A. For this function to make sense, the source geometries must both be of the same coordinate projection, having the same SRID. ST_Contains is the inverse of ST_Within. So ST_Contains(A,B) implies ST_Within(B,A) except in the case of invalid geometries where the result is always false regardless or not defined.
Eseguito dal modulo GEOS
Enhanced: 2.3.0 Enhancement to PIP short-circuit extended to support MultiPoints with few points. Prior versions only supported point in polygon.
Do not call with a |
Do not use this function with invalid geometries. You will get unexpected results. |
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. To avoid index use, use the function _ST_Contains.
NOTE: this is the "allowable" version that returns a boolean, not an integer.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2 // s2.1.13.3 - same as within(geometry B, geometry A)
This method implements the SQL/MM specification. SQL-MM 3: 5.1.31
There are certain subtleties to ST_Contains and ST_Within that are not intuitively obvious. For details check out Subtleties of OGC Covers, Contains, Within
The ST_Contains
predicate returns TRUE
in all the following illustrations.
The ST_Contains
predicate returns FALSE
in all the following illustrations.
-- A circle within a circle SELECT ST_Contains(smallc, bigc) As smallcontainsbig, ST_Contains(bigc,smallc) As bigcontainssmall, ST_Contains(bigc, ST_Union(smallc, bigc)) as bigcontainsunion, ST_Equals(bigc, ST_Union(smallc, bigc)) as bigisunion, ST_Covers(bigc, ST_ExteriorRing(bigc)) As bigcoversexterior, ST_Contains(bigc, ST_ExteriorRing(bigc)) As bigcontainsexterior FROM (SELECT ST_Buffer(ST_GeomFromText('POINT(1 2)'), 10) As smallc, ST_Buffer(ST_GeomFromText('POINT(1 2)'), 20) As bigc) As foo; -- Result smallcontainsbig | bigcontainssmall | bigcontainsunion | bigisunion | bigcoversexterior | bigcontainsexterior ------------------+------------------+------------------+------------+-------------------+--------------------- f | t | t | t | t | f -- Example demonstrating difference between contains and contains properly SELECT ST_GeometryType(geomA) As geomtype, ST_Contains(geomA,geomA) AS acontainsa, ST_ContainsProperly(geomA, geomA) AS acontainspropa, ST_Contains(geomA, ST_Boundary(geomA)) As acontainsba, ST_ContainsProperly(geomA, ST_Boundary(geomA)) As acontainspropba FROM (VALUES ( ST_Buffer(ST_Point(1,1), 5,1) ), ( ST_MakeLine(ST_Point(1,1), ST_Point(-1,-1) ) ), ( ST_Point(1,1) ) ) As foo(geomA); geomtype | acontainsa | acontainspropa | acontainsba | acontainspropba --------------+------------+----------------+-------------+----------------- ST_Polygon | t | f | f | f ST_LineString | t | f | f | f ST_Point | t | t | f | f
ST_ContainsProperly — Returns true if B intersects the interior of A but not the boundary (or exterior). A does not contain properly itself, but does contain itself.
boolean ST_ContainsProperly(
geometry geomA, geometry geomB)
;
Returns true if B intersects the interior of A but not the boundary (or exterior).
A does not contain properly itself, but does contain itself.
Every point of the other geometry is a point of this geometry's interior. The DE-9IM Intersection Matrix for the two geometries matches [T**FF*FF*] used in ST_Relate
From JTS docs slightly reworded: The advantage to using this predicate over ST_Contains and ST_Intersects is that it can be computed efficiently, with no need to compute topology at individual points. An example use case for this predicate is computing the intersections of a set of geometries with a large polygonal geometry. Since intersection is a fairly slow operation, it can be more efficient to use containsProperly to filter out test geometries which lie wholly inside the area. In these cases the intersection is known a priori to be exactly the original test geometry. |
Availability: 1.4.0 - requires GEOS >= 3.1.0.
Do not call with a |
Do not use this function with invalid geometries. You will get unexpected results. |
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. To avoid index use, use the function _ST_ContainsProperly.
--a circle within a circle SELECT ST_ContainsProperly(smallc, bigc) As smallcontainspropbig, ST_ContainsProperly(bigc,smallc) As bigcontainspropsmall, ST_ContainsProperly(bigc, ST_Union(smallc, bigc)) as bigcontainspropunion, ST_Equals(bigc, ST_Union(smallc, bigc)) as bigisunion, ST_Covers(bigc, ST_ExteriorRing(bigc)) As bigcoversexterior, ST_ContainsProperly(bigc, ST_ExteriorRing(bigc)) As bigcontainsexterior FROM (SELECT ST_Buffer(ST_GeomFromText('POINT(1 2)'), 10) As smallc, ST_Buffer(ST_GeomFromText('POINT(1 2)'), 20) As bigc) As foo; --Result smallcontainspropbig | bigcontainspropsmall | bigcontainspropunion | bigisunion | bigcoversexterior | bigcontainsexterior ------------------+------------------+------------------+------------+-------------------+--------------------- f | t | f | t | t | f --example demonstrating difference between contains and contains properly SELECT ST_GeometryType(geomA) As geomtype, ST_Contains(geomA,geomA) AS acontainsa, ST_ContainsProperly(geomA, geomA) AS acontainspropa, ST_Contains(geomA, ST_Boundary(geomA)) As acontainsba, ST_ContainsProperly(geomA, ST_Boundary(geomA)) As acontainspropba FROM (VALUES ( ST_Buffer(ST_Point(1,1), 5,1) ), ( ST_MakeLine(ST_Point(1,1), ST_Point(-1,-1) ) ), ( ST_Point(1,1) ) ) As foo(geomA); geomtype | acontainsa | acontainspropa | acontainsba | acontainspropba --------------+------------+----------------+-------------+----------------- ST_Polygon | t | f | f | f ST_LineString | t | f | f | f ST_Point | t | t | f | f
ST_GeometryType, ST_Boundary, ST_Contains, ST_Covers, ST_CoveredBy, ST_Equals, ST_Relate, ST_Within
ST_Covers — Returns 1 (TRUE) if no point in Geometry B is outside Geometry A
boolean ST_Covers(
geometry geomA, geometry geomB)
;
boolean ST_Covers(
geography geogpolyA, geography geogpointB)
;
Returns 1 (TRUE) if no point in Geometry/Geography B is outside Geometry/Geography A
Eseguito dal modulo GEOS
Do not call with a |
For geography only Polygon covers point is supported. |
Do not use this function with invalid geometries. You will get unexpected results. |
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. To avoid index use, use the function _ST_Covers.
Enhanced: 2.3.0 Enhancement to PIP short-circuit for geometry extended to support MultiPoints with few points. Prior versions only supported point in polygon.
Availability: 1.5 - support for geography was introduced.
Availability: 1.2.2 - requires GEOS >= 3.0
NOTE: this is the "allowable" version that returns a boolean, not an integer.
Non è uno standard OGC, ma è presente anche in Oracle.
There are certain subtleties to ST_Contains and ST_Within that are not intuitively obvious. For details check out Subtleties of OGC Covers, Contains, Within
Geometry example
--a circle covering a circle SELECT ST_Covers(smallc,smallc) As smallinsmall, ST_Covers(smallc, bigc) As smallcoversbig, ST_Covers(bigc, ST_ExteriorRing(bigc)) As bigcoversexterior, ST_Contains(bigc, ST_ExteriorRing(bigc)) As bigcontainsexterior FROM (SELECT ST_Buffer(ST_GeomFromText('POINT(1 2)'), 10) As smallc, ST_Buffer(ST_GeomFromText('POINT(1 2)'), 20) As bigc) As foo; --Result smallinsmall | smallcoversbig | bigcoversexterior | bigcontainsexterior --------------+----------------+-------------------+--------------------- t | f | t | f (1 row)
Geeography Example
-- a point with a 300 meter buffer compared to a point, a point and its 10 meter buffer SELECT ST_Covers(geog_poly, geog_pt) As poly_covers_pt, ST_Covers(ST_Buffer(geog_pt,10), geog_pt) As buff_10m_covers_cent FROM (SELECT ST_Buffer(ST_GeogFromText('SRID=4326;POINT(-99.327 31.4821)'), 300) As geog_poly, ST_GeogFromText('SRID=4326;POINT(-99.33 31.483)') As geog_pt ) As foo; poly_covers_pt | buff_10m_covers_cent ----------------+------------------ f | t
ST_CoveredBy — Returns 1 (TRUE) if no point in Geometry/Geography A is outside Geometry/Geography B
boolean ST_CoveredBy(
geometry geomA, geometry geomB)
;
boolean ST_CoveredBy(
geography geogA, geography geogB)
;
Returns 1 (TRUE) if no point in Geometry/Geography A is outside Geometry/Geography B
Eseguito dal modulo GEOS
Do not call with a |
Do not use this function with invalid geometries. You will get unexpected results. |
Availability: 1.2.2 - requires GEOS >= 3.0
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. To avoid index use, use the function _ST_CoveredBy.
NOTE: this is the "allowable" version that returns a boolean, not an integer.
Non è uno standard OGC, ma è presente anche in Oracle.
There are certain subtleties to ST_Contains and ST_Within that are not intuitively obvious. For details check out Subtleties of OGC Covers, Contains, Within
--a circle coveredby a circle SELECT ST_CoveredBy(smallc,smallc) As smallinsmall, ST_CoveredBy(smallc, bigc) As smallcoveredbybig, ST_CoveredBy(ST_ExteriorRing(bigc), bigc) As exteriorcoveredbybig, ST_Within(ST_ExteriorRing(bigc),bigc) As exeriorwithinbig FROM (SELECT ST_Buffer(ST_GeomFromText('POINT(1 2)'), 10) As smallc, ST_Buffer(ST_GeomFromText('POINT(1 2)'), 20) As bigc) As foo; --Result smallinsmall | smallcoveredbybig | exteriorcoveredbybig | exeriorwithinbig --------------+-------------------+----------------------+------------------ t | t | t | f (1 row)
ST_Crosses — Returns TRUE
if the supplied geometries have some, but not all, interior points in common.
boolean ST_Crosses(
geometry g1, geometry g2)
;
ST_Crosses
takes two geometry objects and returns TRUE
if their intersection "spatially cross", that is, the geometries have some, but not all interior points in common. The intersection of the interiors of the geometries must not be the empty set and must have a dimensionality less than the maximum dimension of the two input geometries. Additionally, the intersection of the two geometries must not equal either of the source geometries. Otherwise, it returns FALSE
.
In mathematical terms, this is expressed as:
TODO: Insert appropriate MathML markup here or use a gif. Simple HTML markup does not work well in both IE and Firefox.
The DE-9IM Intersection Matrix for the two geometries is:
T*T****** (per casi Point/Line, Point/Area, e Line/Area)
T*****T** (for Line/Point, Area/Point, and Area/Line situations)
0******** (for Line/Line situations)
For any other combination of dimensions this predicate returns false.
The OpenGIS Simple Features Specification defines this predicate only for Point/Line, Point/Area, Line/Line, and Line/Area situations. JTS / GEOS extends the definition to apply to Line/Point, Area/Point and Area/Line situations as well. This makes the relation symmetric.
Do not call with a |
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.13.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.29
The following illustrations all return TRUE
.
Consider a situation where a user has two tables: a table of roads and a table of highways.
CREATE TABLE roads ( id serial NOT NULL, the_geom geometry, CONSTRAINT roads_pkey PRIMARY KEY (road_id) );
|
CREATE TABLE highways ( id serial NOT NULL, the_gem geometry, CONSTRAINT roads_pkey PRIMARY KEY (road_id) );
|
To determine a list of roads that cross a highway, use a query similiar to:
SELECT roads.id FROM roads, highways WHERE ST_Crosses(roads.the_geom, highways.the_geom);
ST_LineCrossingDirection — Given 2 linestrings, returns a number between -3 and 3 denoting what kind of crossing behavior. 0 is no crossing.
integer ST_LineCrossingDirection(
geometry linestringA, geometry linestringB)
;
Given 2 linestrings, returns a number between -3 and 3 denoting what kind of crossing behavior. 0 is no crossing. This is only supported for LINESTRING
Definition of integer constants is as follows:
0: LINE NO CROSS
-1: LINE CROSS LEFT
1: LINE CROSS RIGHT
-2: LINE MULTICROSS END LEFT
2: LINE MULTICROSS END RIGHT
-3: LINE MULTICROSS END SAME FIRST LEFT
3: LINE MULTICROSS END SAME FIRST RIGHT
Disponibilità: 1.4.0
SELECT ST_LineCrossingDirection(foo.line1, foo.line2) As l1_cross_l2 , ST_LineCrossingDirection(foo.line2, foo.line1) As l2_cross_l1 FROM ( SELECT ST_GeomFromText('LINESTRING(25 169,89 114,40 70,86 43)') As line1, ST_GeomFromText('LINESTRING(171 154,20 140,71 74,161 53)') As line2 ) As foo; l1_cross_l2 | l2_cross_l1 -------------+------------- 3 | -3
|
SELECT ST_LineCrossingDirection(foo.line1, foo.line2) As l1_cross_l2 , ST_LineCrossingDirection(foo.line2, foo.line1) As l2_cross_l1 FROM ( SELECT ST_GeomFromText('LINESTRING(25 169,89 114,40 70,86 43)') As line1, ST_GeomFromText('LINESTRING (171 154, 20 140, 71 74, 2.99 90.16)') As line2 ) As foo; l1_cross_l2 | l2_cross_l1 -------------+------------- 2 | -2
|
SELECT ST_LineCrossingDirection(foo.line1, foo.line2) As l1_cross_l2 , ST_LineCrossingDirection(foo.line2, foo.line1) As l2_cross_l1 FROM ( SELECT ST_GeomFromText('LINESTRING(25 169,89 114,40 70,86 43)') As line1, ST_GeomFromText('LINESTRING (20 140, 71 74, 161 53)') As line2 ) As foo; l1_cross_l2 | l2_cross_l1 -------------+------------- -1 | 1
|
SELECT ST_LineCrossingDirection(foo.line1, foo.line2) As l1_cross_l2 , ST_LineCrossingDirection(foo.line2, foo.line1) As l2_cross_l1 FROM (SELECT ST_GeomFromText('LINESTRING(25 169,89 114,40 70,86 43)') As line1, ST_GeomFromText('LINESTRING(2.99 90.16,71 74,20 140,171 154)') As line2 ) As foo; l1_cross_l2 | l2_cross_l1 -------------+------------- -2 | 2
|
SELECT s1.gid, s2.gid, ST_LineCrossingDirection(s1.the_geom, s2.the_geom) FROM streets s1 CROSS JOIN streets s2 ON (s1.gid != s2.gid AND s1.the_geom && s2.the_geom ) WHERE ST_CrossingDirection(s1.the_geom, s2.the_geom) > 0;
ST_Disjoint — Returns TRUE if the Geometries do not "spatially intersect" - if they do not share any space together.
boolean ST_Disjoint(
geometry A , geometry B )
;
Overlaps, Touches, Within all imply geometries are not spatially disjoint. If any of the aforementioned returns true, then the geometries are not spatially disjoint. Disjoint implies false for spatial intersection.
Do not call with a |
Eseguito dal modulo GEOS
This function call does not use indexes |
NOTE: this is the "allowable" version that returns a boolean, not an integer. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2 //s2.1.13.3 - a.Relate(b, 'FF*FF****')
This method implements the SQL/MM specification. SQL-MM 3: 5.1.26
ST_Distance — For geometry type Returns the 2D Cartesian distance between two geometries in projected units (based on spatial ref). For geography type defaults to return minimum geodesic distance between two geographies in meters.
float ST_Distance(
geometry g1, geometry g2)
;
float ST_Distance(
geography gg1, geography gg2)
;
float ST_Distance(
geography gg1, geography gg2, boolean use_spheroid)
;
For geometry type returns the minimum 2D Cartesian distance between two geometries in projected units (spatial ref units). For geography type defaults to return the minimum geodesic distance between two geographies in meters. If use_spheroid is false, a faster sphere calculation is used instead of a spheroid.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 5.1.23
This method supports Circular Strings and Curves
This method is also provided by SFCGAL backend.
Availability: 1.5.0 geography support was introduced in 1.5. Speed improvements for planar to better handle large or many vertex geometries
Enhanced: 2.1.0 improved speed for geography. See Making Geography faster for details.
Enhanced: 2.1.0 - support for curved geometries was introduced.
Enhanced: 2.2.0 - measurement on spheroid performed with GeographicLib for improved accuracy and robustness. Requires Proj >= 4.9.0 to take advantage of the new feature.
--Geometry example - units in planar degrees 4326 is WGS 84 long lat unit=degrees SELECT ST_Distance( ST_GeomFromText('POINT(-72.1235 42.3521)',4326), ST_GeomFromText('LINESTRING(-72.1260 42.45, -72.123 42.1546)', 4326) ); st_distance ----------------- 0.00150567726382282 -- Geometry example - units in meters (SRID: 26986 Massachusetts state plane meters) (most accurate for Massachusetts) SELECT ST_Distance( ST_Transform(ST_GeomFromText('POINT(-72.1235 42.3521)',4326),26986), ST_Transform(ST_GeomFromText('LINESTRING(-72.1260 42.45, -72.123 42.1546)', 4326),26986) ); st_distance ----------------- 123.797937878454 -- Geometry example - units in meters (SRID: 2163 US National Atlas Equal area) (least accurate) SELECT ST_Distance( ST_Transform(ST_GeomFromText('POINT(-72.1235 42.3521)',4326),2163), ST_Transform(ST_GeomFromText('LINESTRING(-72.1260 42.45, -72.123 42.1546)', 4326),2163) ); st_distance ------------------ 126.664256056812
-- same as geometry example but note units in meters - use sphere for slightly faster less accurate SELECT ST_Distance(gg1, gg2) As spheroid_dist, ST_Distance(gg1, gg2, false) As sphere_dist FROM (SELECT ST_GeogFromText('SRID=4326;POINT(-72.1235 42.3521)') As gg1, ST_GeogFromText('SRID=4326;LINESTRING(-72.1260 42.45, -72.123 42.1546)') As gg2 ) As foo ; spheroid_dist | sphere_dist ------------------+------------------ 123.802076746848 | 123.475736916397
ST_3DDistance, ST_DWithin, ST_DistanceSphere, ST_DistanceSpheroid, ST_MaxDistance, ST_Transform
ST_HausdorffDistance — Returns the Hausdorff distance between two geometries. Basically a measure of how similar or dissimilar 2 geometries are. Units are in the units of the spatial reference system of the geometries.
float ST_HausdorffDistance(
geometry g1, geometry g2)
;
float ST_HausdorffDistance(
geometry g1, geometry g2, float densifyFrac)
;
Implements algorithm for computing a distance metric which can be thought of as the "Discrete Hausdorff Distance". This is the Hausdorff distance restricted to discrete points for one of the geometries. Wikipedia article on Hausdorff distance Martin Davis note on how Hausdorff Distance calculation was used to prove correctness of the CascadePolygonUnion approach.
When densifyFrac is specified, this function performs a segment densification before computing the discrete hausdorff distance. The densifyFrac parameter sets the fraction by which to densify each segment. Each segment will be split into a number of equal-length subsegments, whose fraction of the total length is closest to the given fraction.
The current implementation supports only vertices as the discrete locations. This could be extended to allow an arbitrary density of points to be used. |
This algorithm is NOT equivalent to the standard Hausdorff distance. However, it computes an approximation that is correct for a large subset of useful cases. One important part of this subset is Linestrings that are roughly parallel to each other, and roughly equal in length. This is a useful metric for line matching. |
Availability: 1.5.0 - requires GEOS >= 3.2.0
For each building, find the parcel that best represents it. First we require the parcel intersect with the geometry. DISTINCT ON guarantees we get each building listed only once, the ORDER BY .. ST_HausdorffDistance gives us a preference of parcel that is most similar to the building.
SELECT DISTINCT ON(buildings.gid) buildings.gid, parcels.parcel_id FROM buildings INNER JOIN parcels ON ST_Intersects(buildings.geom,parcels.geom) ORDER BY buildings.gid, ST_HausdorffDistance(buildings.geom, parcels.geom);
postgis=# SELECT ST_HausdorffDistance( 'LINESTRING (0 0, 2 0)'::geometry, 'MULTIPOINT (0 1, 1 0, 2 1)'::geometry); st_hausdorffdistance ---------------------- 1 (1 row)
postgis=# SELECT st_hausdorffdistance('LINESTRING (130 0, 0 0, 0 150)'::geometry, 'LINESTRING (10 10, 10 150, 130 10)'::geometry, 0.5); st_hausdorffdistance ---------------------- 70 (1 row)
ST_MaxDistance — Returns the 2-dimensional largest distance between two geometries in projected units.
float ST_MaxDistance(
geometry g1, geometry g2)
;
Returns the 2-dimensional maximum distance between two geometries in projected units. If g1 and g2 is the same geometry the function will return the distance between the two vertices most far from each other in that geometry. |
Disponibilità: 1.5.0
Basic furthest distance the point is to any part of the line
postgis=# SELECT ST_MaxDistance('POINT(0 0)'::geometry, 'LINESTRING ( 2 0, 0 2 )'::geometry); st_maxdistance ----------------- 2 (1 row) postgis=# SELECT ST_MaxDistance('POINT(0 0)'::geometry, 'LINESTRING ( 2 2, 2 2 )'::geometry); st_maxdistance ------------------ 2.82842712474619 (1 row)
ST_DistanceSphere — Returns minimum distance in meters between two lon/lat geometries. Uses a spherical earth and radius derived from the spheroid defined by the SRID. Faster than ST_DistanceSpheroid ST_DistanceSpheroid, but less accurate. PostGIS versions prior to 1.5 only implemented for points.
float ST_DistanceSphere(
geometry geomlonlatA, geometry geomlonlatB)
;
Returns minimum distance in meters between two lon/lat points. Uses a spherical earth and radius derived from the spheroid defined by the SRID. Faster than ST_DistanceSpheroid, but less accurate. PostGIS Versions prior to 1.5 only implemented for points.
Availability: 1.5 - support for other geometry types besides points was introduced. Prior versions only work with points.
Changed: 2.2.0 In prior versions this used to be called ST_Distance_Sphere
SELECT round(CAST(ST_DistanceSphere(ST_Centroid(the_geom), ST_GeomFromText('POINT(-118 38)',4326)) As numeric),2) As dist_meters, round(CAST(ST_Distance(ST_Transform(ST_Centroid(the_geom),32611), ST_Transform(ST_GeomFromText('POINT(-118 38)', 4326),32611)) As numeric),2) As dist_utm11_meters, round(CAST(ST_Distance(ST_Centroid(the_geom), ST_GeomFromText('POINT(-118 38)', 4326)) As numeric),5) As dist_degrees, round(CAST(ST_Distance(ST_Transform(the_geom,32611), ST_Transform(ST_GeomFromText('POINT(-118 38)', 4326),32611)) As numeric),2) As min_dist_line_point_meters FROM (SELECT ST_GeomFromText('LINESTRING(-118.584 38.374,-118.583 38.5)', 4326) As the_geom) as foo; dist_meters | dist_utm11_meters | dist_degrees | min_dist_line_point_meters -------------+-------------------+--------------+---------------------------- 70424.47 | 70438.00 | 0.72900 | 65871.18
ST_DistanceSpheroid — Returns the minimum distance between two lon/lat geometries given a particular spheroid. PostGIS versions prior to 1.5 only support points.
float ST_DistanceSpheroid(
geometry geomlonlatA, geometry geomlonlatB, spheroid measurement_spheroid)
;
Returns minimum distance in meters between two lon/lat geometries given a particular spheroid. See the explanation of spheroids given for ST_LengthSpheroid. PostGIS version prior to 1.5 only support points.
This function currently does not look at the SRID of a geometry and will always assume its represented in the coordinates of the passed in spheroid. Prior versions of this function only support points. |
Availability: 1.5 - support for other geometry types besides points was introduced. Prior versions only work with points.
Changed: 2.2.0 In prior versions this used to be called ST_Distance_Spheroid
SELECT round(CAST( ST_DistanceSpheroid(ST_Centroid(the_geom), ST_GeomFromText('POINT(-118 38)',4326), 'SPHEROID["WGS 84",6378137,298.257223563]') As numeric),2) As dist_meters_spheroid, round(CAST(ST_DistanceSphere(ST_Centroid(the_geom), ST_GeomFromText('POINT(-118 38)',4326)) As numeric),2) As dist_meters_sphere, round(CAST(ST_Distance(ST_Transform(ST_Centroid(the_geom),32611), ST_Transform(ST_GeomFromText('POINT(-118 38)', 4326),32611)) As numeric),2) As dist_utm11_meters FROM (SELECT ST_GeomFromText('LINESTRING(-118.584 38.374,-118.583 38.5)', 4326) As the_geom) as foo; dist_meters_spheroid | dist_meters_sphere | dist_utm11_meters ----------------------+--------------------+------------------- 70454.92 | 70424.47 | 70438.00
ST_DFullyWithin — Returns true if all of the geometries are within the specified distance of one another
boolean ST_DFullyWithin(
geometry g1, geometry g2, double precision distance)
;
Returns true if the geometries is fully within the specified distance of one another. The distance is specified in units defined by the spatial reference system of the geometries. For this function to make sense, the source geometries must both be of the same coordinate projection, having the same SRID.
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. |
Disponibilità: 1.5.0
postgis=# SELECT ST_DFullyWithin(geom_a, geom_b, 10) as DFullyWithin10, ST_DWithin(geom_a, geom_b, 10) as DWithin10, ST_DFullyWithin(geom_a, geom_b, 20) as DFullyWithin20 from (select ST_GeomFromText('POINT(1 1)') as geom_a,ST_GeomFromText('LINESTRING(1 5, 2 7, 1 9, 14 12)') as geom_b) t1; ----------------- DFullyWithin10 | DWithin10 | DFullyWithin20 | ---------------+----------+---------------+ f | t | t |
ST_DWithin — Returns true if the geometries are within the specified distance of one another. For geometry units are in those of spatial reference and For geography units are in meters and measurement is defaulted to use_spheroid=true (measure around spheroid), for faster check, use_spheroid=false to measure along sphere.
boolean ST_DWithin(
geometry g1, geometry g2, double precision distance_of_srid)
;
boolean ST_DWithin(
geography gg1, geography gg2, double precision distance_meters)
;
boolean ST_DWithin(
geography gg1, geography gg2, double precision distance_meters, boolean use_spheroid)
;
Returns true if the geometries are within the specified distance of one another.
For Geometries: The distance is specified in units defined by the spatial reference system of the geometries. For this function to make sense, the source geometries must both be of the same coordinate projection, having the same SRID.
For geography units are in meters and measurement is defaulted to use_spheroid=true, for faster check, use_spheroid=false to measure along sphere.
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. |
Prior to 1.3, ST_Expand was commonly used in conjunction with && and ST_Distance to achieve the same effect and in pre-1.3.4 this function was basically short-hand for that construct. From 1.3.4, ST_DWithin uses a more short-circuit distance function which should make it more efficient than prior versions for larger buffer regions. |
Use ST_3DDWithin if you have 3D geometries. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
Availability: 1.5.0 support for geography was introduced
Enhanced: 2.1.0 improved speed for geography. See Making Geography faster for details.
Enhanced: 2.1.0 support for curved geometries was introduced.
--Find the nearest hospital to each school --that is within 3000 units of the school. -- We do an ST_DWithin search to utilize indexes to limit our search list -- that the non-indexable ST_Distance needs to process --If the units of the spatial reference is meters then units would be meters SELECT DISTINCT ON (s.gid) s.gid, s.school_name, s.the_geom, h.hospital_name FROM schools s LEFT JOIN hospitals h ON ST_DWithin(s.the_geom, h.the_geom, 3000) ORDER BY s.gid, ST_Distance(s.the_geom, h.the_geom); --The schools with no close hospitals --Find all schools with no hospital within 3000 units --away from the school. Units is in units of spatial ref (e.g. meters, feet, degrees) SELECT s.gid, s.school_name FROM schools s LEFT JOIN hospitals h ON ST_DWithin(s.the_geom, h.the_geom, 3000) WHERE h.gid IS NULL;
ST_Equals — Returns true if the given geometries represent the same geometry. Directionality is ignored.
boolean ST_Equals(
geometry A, geometry B)
;
Returns TRUE if the given Geometries are "spatially equal". Use this for a 'better' answer than '='. Note by spatially equal we mean ST_Within(A,B) = true and ST_Within(B,A) = true and also mean ordering of points can be different but represent the same geometry structure. To verify the order of points is consistent, use ST_OrderingEquals (it must be noted ST_OrderingEquals is a little more stringent than simply verifying order of points are the same).
This function will return false if either geometry is invalid even if they are binary equal. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2
This method implements the SQL/MM specification. SQL-MM 3: 5.1.24
ST_HasArc — Returns true if a geometry or geometry collection contains a circular string
boolean ST_HasArc(
geometry geomA)
;
Returns true if a geometry or geometry collection contains a circular string
Availability: 1.2.3?
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
ST_Intersects — Returns TRUE if the Geometries/Geography "spatially intersect in 2D" - (share any portion of space) and FALSE if they don't (they are Disjoint). For geography -- tolerance is 0.00001 meters (so any points that close are considered to intersect)
boolean ST_Intersects(
geometry geomA , geometry geomB )
;
boolean ST_Intersects(
geography geogA , geography geogB )
;
If a geometry or geography shares any portion of space then they intersect. For geography -- tolerance is 0.00001 meters (so any points that are close are considered to intersect)
Overlaps, Touches, Within all imply spatial intersection. If any of the aforementioned returns true, then the geometries also spatially intersect. Disjoint implies false for spatial intersection.
Do not call with a |
Enhanced: 2.3.0 Enhancement to PIP short-circuit extended to support MultiPoints with few points. Prior versions only supported point in polygon.
Performed by the GEOS module (for geometry), geography is native
Availability: 1.5 support for geography was introduced.
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. |
For geography, this function has a distance tolerance of about 0.00001 meters and uses the sphere rather than spheroid calculation. |
NOTE: this is the "allowable" version that returns a boolean, not an integer. |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2 //s2.1.13.3 - ST_Intersects(g1, g2 ) --> Not (ST_Disjoint(g1, g2 ))
This method implements the SQL/MM specification. SQL-MM 3: 5.1.27
This method is also provided by SFCGAL backend.
SELECT ST_Intersects('POINT(0 0)'::geometry, 'LINESTRING ( 2 0, 0 2 )'::geometry); st_intersects --------------- f (1 row) SELECT ST_Intersects('POINT(0 0)'::geometry, 'LINESTRING ( 0 0, 0 2 )'::geometry); st_intersects --------------- t (1 row)
ST_Length — Returns the 2D length of the geometry if it is a LineString or MultiLineString. geometry are in units of spatial reference and geography are in meters (default spheroid)
float ST_Length(
geometry a_2dlinestring)
;
float ST_Length(
geography geog, boolean use_spheroid=true)
;
For geometry: Returns the 2D Cartesian length of the geometry if it is a LineString, MultiLineString, ST_Curve, ST_MultiCurve. 0 is returned for areal geometries. For areal geometries use ST_Perimeter. For geometry types, units for length measures are specified by the spatial reference system of the geometry.
For geography types, the calculations are performed using the inverse geodesic problem, where length units are in meters. If PostGIS is compiled with PROJ version 4.8.0 or later, the spheroid is specified by the SRID, otherwise it is exclusive to WGS84. If use_spheroid=false
, then calculations will approximate a sphere instead of a spheroid.
Currently for geometry this is an alias for ST_Length2D, but this may change to support higher dimensions.
Changed: 2.0.0 Breaking change -- in prior versions applying this to a MULTI/POLYGON of type geography would give you the perimeter of the POLYGON/MULTIPOLYGON. In 2.0.0 this was changed to return 0 to be in line with geometry behavior. Please use ST_Perimeter if you want the perimeter of a polygon |
For geography measurement defaults spheroid measurement. To use the faster less accurate sphere use ST_Length(gg,false); |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.5.1
This method implements the SQL/MM specification. SQL-MM 3: 7.1.2, 9.3.4
Availability: 1.5.0 geography support was introduced in 1.5.
This method is also provided by SFCGAL backend.
Return length in feet for line string. Note this is in feet because EPSG:2249 is Massachusetts State Plane Feet
SELECT ST_Length(ST_GeomFromText('LINESTRING(743238 2967416,743238 2967450,743265 2967450, 743265.625 2967416,743238 2967416)',2249)); st_length --------- 122.630744000095 --Transforming WGS 84 LineString to Massachusetts state plane meters SELECT ST_Length( ST_Transform( ST_GeomFromEWKT('SRID=4326;LINESTRING(-72.1260 42.45, -72.1240 42.45666, -72.123 42.1546)'), 26986 ) ); st_length --------- 34309.4563576191
Return length of WGS 84 geography line
-- default calculation is using a sphere rather than spheroid SELECT ST_Length(the_geog) As length_spheroid, ST_Length(the_geog,false) As length_sphere FROM (SELECT ST_GeographyFromText( 'SRID=4326;LINESTRING(-72.1260 42.45, -72.1240 42.45666, -72.123 42.1546)') As the_geog) As foo; length_spheroid | length_sphere ------------------+------------------ 34310.5703627288 | 34346.2060960742
ST_Length2D — Returns the 2-dimensional length of the geometry if it is a linestring or multi-linestring. This is an alias for ST_Length
float ST_Length2D(
geometry a_2dlinestring)
;
ST_3DLength — Returns the 3-dimensional or 2-dimensional length of the geometry if it is a linestring or multi-linestring.
float ST_3DLength(
geometry a_3dlinestring)
;
Returns the 3-dimensional or 2-dimensional length of the geometry if it is a linestring or multi-linestring. For 2-d lines it will just return the 2-d length (same as ST_Length and ST_Length2D)
This function supports 3d and will not drop the z-index.
Changed: 2.0.0 In prior versions this used to be called ST_Length3D
Return length in feet for a 3D cable. Note this is in feet because EPSG:2249 is Massachusetts State Plane Feet
SELECT ST_3DLength(ST_GeomFromText('LINESTRING(743238 2967416 1,743238 2967450 1,743265 2967450 3, 743265.625 2967416 3,743238 2967416 3)',2249)); ST_3DLength ----------- 122.704716741457
ST_LengthSpheroid — Calculates the 2D or 3D length of a linestring/multilinestring on an ellipsoid. This is useful if the coordinates of the geometry are in longitude/latitude and a length is desired without reprojection.
float ST_LengthSpheroid(
geometry a_linestring, spheroid a_spheroid)
;
Calculates the length of a geometry on an ellipsoid. This is useful if the coordinates of the geometry are in longitude/latitude and a length is desired without reprojection. The ellipsoid is a separate database type and can be constructed as follows:
SPHEROID[<NAME>,<SEMI-MAJOR
AXIS>,<INVERSE FLATTENING>]
SPHEROID["GRS_1980",6378137,298.257222101]
Will return 0 for anything that is not a MULTILINESTRING or LINESTRING |
Availability: 1.2.2
Changed: 2.2.0 In prior versions this used to be called ST_Length_Spheroid and used to have a ST_3DLength_Spheroid alias
This function supports 3d and will not drop the z-index.
SELECT ST_LengthSpheroid( geometry_column, 'SPHEROID["GRS_1980",6378137,298.257222101]' ) FROM geometry_table; SELECT ST_LengthSpheroid( the_geom, sph_m ) As tot_len, ST_LengthSpheroid(ST_GeometryN(the_geom,1), sph_m) As len_line1, ST_LengthSpheroid(ST_GeometryN(the_geom,2), sph_m) As len_line2 FROM (SELECT ST_GeomFromText('MULTILINESTRING((-118.584 38.374,-118.583 38.5), (-71.05957 42.3589 , -71.061 43))') As the_geom, CAST('SPHEROID["GRS_1980",6378137,298.257222101]' As spheroid) As sph_m) as foo; tot_len | len_line1 | len_line2 ------------------+------------------+------------------ 85204.5207562955 | 13986.8725229309 | 71217.6482333646 --3D SELECT ST_LengthSpheroid( the_geom, sph_m ) As tot_len, ST_LengthSpheroid(ST_GeometryN(the_geom,1), sph_m) As len_line1, ST_LengthSpheroid(ST_GeometryN(the_geom,2), sph_m) As len_line2 FROM (SELECT ST_GeomFromEWKT('MULTILINESTRING((-118.584 38.374 20,-118.583 38.5 30), (-71.05957 42.3589 75, -71.061 43 90))') As the_geom, CAST('SPHEROID["GRS_1980",6378137,298.257222101]' As spheroid) As sph_m) as foo; tot_len | len_line1 | len_line2 ------------------+-----------------+------------------ 85204.5259107402 | 13986.876097711 | 71217.6498130292
ST_Length2D_Spheroid — Calculates the 2D length of a linestring/multilinestring on an ellipsoid. This is useful if the coordinates of the geometry are in longitude/latitude and a length is desired without reprojection.
float ST_Length2D_Spheroid(
geometry a_linestring, spheroid a_spheroid)
;
Calculates the 2D length of a geometry on an ellipsoid. This is useful if the coordinates of the geometry are in longitude/latitude and a length is desired without reprojection. The ellipsoid is a separate database type and can be constructed as follows:
SPHEROID[<NAME>,<SEMI-MAJOR
AXIS>,<INVERSE FLATTENING>]
SPHEROID["GRS_1980",6378137,298.257222101]
Will return 0 for anything that is not a MULTILINESTRING or LINESTRING |
This is much like ST_LengthSpheroid except it will throw away the Z coordinate in calculations. |
SELECT ST_Length2D_Spheroid( geometry_column, 'SPHEROID["GRS_1980",6378137,298.257222101]' ) FROM geometry_table; SELECT ST_Length2D_Spheroid( the_geom, sph_m ) As tot_len, ST_Length2D_Spheroid(ST_GeometryN(the_geom,1), sph_m) As len_line1, ST_Length2D_Spheroid(ST_GeometryN(the_geom,2), sph_m) As len_line2 FROM (SELECT ST_GeomFromText('MULTILINESTRING((-118.584 38.374,-118.583 38.5), (-71.05957 42.3589 , -71.061 43))') As the_geom, CAST('SPHEROID["GRS_1980",6378137,298.257222101]' As spheroid) As sph_m) as foo; tot_len | len_line1 | len_line2 ------------------+------------------+------------------ 85204.5207562955 | 13986.8725229309 | 71217.6482333646 --3D Observe same answer SELECT ST_Length2D_Spheroid( the_geom, sph_m ) As tot_len, ST_Length2D_Spheroid(ST_GeometryN(the_geom,1), sph_m) As len_line1, ST_Length2D_Spheroid(ST_GeometryN(the_geom,2), sph_m) As len_line2 FROM (SELECT ST_GeomFromEWKT('MULTILINESTRING((-118.584 38.374 20,-118.583 38.5 30), (-71.05957 42.3589 75, -71.061 43 90))') As the_geom, CAST('SPHEROID["GRS_1980",6378137,298.257222101]' As spheroid) As sph_m) as foo; tot_len | len_line1 | len_line2 ------------------+------------------+------------------ 85204.5207562955 | 13986.8725229309 | 71217.6482333646
ST_LongestLine — Returns the 2-dimensional longest line points of two geometries. The function will only return the first longest line if more than one, that the function finds. The line returned will always start in g1 and end in g2. The length of the line this function returns will always be the same as st_maxdistance returns for g1 and g2.
geometry ST_LongestLine(
geometry g1, geometry g2)
;
Returns the 2-dimensional longest line between the points of two geometries.
Disponibilità: 1.5.0
SELECT ST_AsText( ST_LongestLine('POINT(100 100)'::geometry, 'LINESTRING (20 80, 98 190, 110 180, 50 75 )'::geometry) ) As lline; lline ----------------- LINESTRING(100 100,98 190)
|
SELECT ST_AsText( ST_LongestLine( ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) ) ) As llinewkt; lline ----------------- LINESTRING(20 40,121.111404660392 186.629392246051)
|
SELECT ST_AsText(ST_LongestLine(c.the_geom, c.the_geom)) As llinewkt, ST_MaxDistance(c.the_geom,c.the_geom) As max_dist, ST_Length(ST_LongestLine(c.the_geom, c.the_geom)) As lenll FROM (SELECT ST_BuildArea(ST_Collect(the_geom)) As the_geom FROM (SELECT ST_Translate(ST_SnapToGrid(ST_Buffer(ST_Point(50 ,generate_series(50,190, 50) ),40, 'quad_segs=2'),1), x, 0) As the_geom FROM generate_series(1,100,50) As x) AS foo ) As c; llinewkt | max_dist | lenll ---------------------------+------------------+------------------ LINESTRING(23 22,129 178) | 188.605408193933 | 188.605408193933
|
ST_OrderingEquals — Returns true if the given geometries represent the same geometry and points are in the same directional order.
boolean ST_OrderingEquals(
geometry A, geometry B)
;
ST_OrderingEquals compares two geometries and returns t (TRUE) if the geometries are equal and the coordinates are in the same order; otherwise it returns f (FALSE).
This function is implemented as per the ArcSDE SQL specification rather than SQL-MM. http://edndoc.esri.com/arcsde/9.1/sql_api/sqlapi3.htm#ST_OrderingEquals |
This method implements the SQL/MM specification. SQL-MM 3: 5.1.43
SELECT ST_OrderingEquals(ST_GeomFromText('LINESTRING(0 0, 10 10)'), ST_GeomFromText('LINESTRING(0 0, 5 5, 10 10)')); st_orderingequals ----------- f (1 row) SELECT ST_OrderingEquals(ST_GeomFromText('LINESTRING(0 0, 10 10)'), ST_GeomFromText('LINESTRING(0 0, 0 0, 10 10)')); st_orderingequals ----------- t (1 row) SELECT ST_OrderingEquals(ST_Reverse(ST_GeomFromText('LINESTRING(0 0, 10 10)')), ST_GeomFromText('LINESTRING(0 0, 0 0, 10 10)')); st_orderingequals ----------- f (1 row)
ST_Overlaps — Returns TRUE if the Geometries share space, are of the same dimension, but are not completely contained by each other.
boolean ST_Overlaps(
geometry A, geometry B)
;
Returns TRUE if the Geometries "spatially overlap". By that we mean they intersect, but one does not completely contain another.
Eseguito dal modulo GEOS
Do not call with a GeometryCollection as an argument |
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. To avoid index use, use the function _ST_Overlaps.
NOTE: this is the "allowable" version that returns a boolean, not an integer.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2 // s2.1.13.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.32
The following illustrations all return TRUE
.
--a point on a line is contained by the line and is of a lower dimension, and therefore does not overlap the line nor crosses SELECT ST_Overlaps(a,b) As a_overlap_b, ST_Crosses(a,b) As a_crosses_b, ST_Intersects(a, b) As a_intersects_b, ST_Contains(b,a) As b_contains_a FROM (SELECT ST_GeomFromText('POINT(1 0.5)') As a, ST_GeomFromText('LINESTRING(1 0, 1 1, 3 5)') As b) As foo a_overlap_b | a_crosses_b | a_intersects_b | b_contains_a ------------+-------------+----------------+-------------- f | f | t | t --a line that is partly contained by circle, but not fully is defined as intersecting and crossing, -- but since of different dimension it does not overlap SELECT ST_Overlaps(a,b) As a_overlap_b, ST_Crosses(a,b) As a_crosses_b, ST_Intersects(a, b) As a_intersects_b, ST_Contains(a,b) As a_contains_b FROM (SELECT ST_Buffer(ST_GeomFromText('POINT(1 0.5)'), 3) As a, ST_GeomFromText('LINESTRING(1 0, 1 1, 3 5)') As b) As foo; a_overlap_b | a_crosses_b | a_intersects_b | a_contains_b -------------+-------------+----------------+-------------- f | t | t | f -- a 2-dimensional bent hot dog (aka buffered line string) that intersects a circle, -- but is not fully contained by the circle is defined as overlapping since they are of the same dimension, -- but it does not cross, because the intersection of the 2 is of the same dimension -- as the maximum dimension of the 2 SELECT ST_Overlaps(a,b) As a_overlap_b, ST_Crosses(a,b) As a_crosses_b, ST_Intersects(a, b) As a_intersects_b, ST_Contains(b,a) As b_contains_a, ST_Dimension(a) As dim_a, ST_Dimension(b) as dim_b, ST_Dimension(ST_Intersection(a,b)) As dima_intersection_b FROM (SELECT ST_Buffer(ST_GeomFromText('POINT(1 0.5)'), 3) As a, ST_Buffer(ST_GeomFromText('LINESTRING(1 0, 1 1, 3 5)'),0.5) As b) As foo; a_overlap_b | a_crosses_b | a_intersects_b | b_contains_a | dim_a | dim_b | dima_intersection_b -------------+-------------+----------------+--------------+-------+-------+--------------------- t | f | t | f | 2 | 2 | 2
ST_Perimeter — Return the length measurement of the boundary of an ST_Surface or ST_MultiSurface geometry or geography. (Polygon, MultiPolygon). geometry measurement is in units of spatial reference and geography is in meters.
float ST_Perimeter(
geometry g1)
;
float ST_Perimeter(
geography geog, boolean use_spheroid=true)
;
Returns the 2D perimeter of the geometry/geography if it is a ST_Surface, ST_MultiSurface (Polygon, MultiPolygon). 0 is returned for non-areal geometries. For linear geometries use ST_Length. For geometry types, units for perimeter measures are specified by the spatial reference system of the geometry.
For geography types, the calculations are performed using the inverse geodesic problem, where perimeter units are in meters. If PostGIS is compiled with PROJ version 4.8.0 or later, the spheroid is specified by the SRID, otherwise it is exclusive to WGS84. If use_spheroid=false
, then calculations will approximate a sphere instead of a spheroid.
Currently this is an alias for ST_Perimeter2D, but this may change to support higher dimensions.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.5.1
This method implements the SQL/MM specification. SQL-MM 3: 8.1.3, 9.5.4
Availability 2.0.0: Support for geography was introduced
Return perimeter in feet for Polygon and MultiPolygon. Note this is in feet because EPSG:2249 is Massachusetts State Plane Feet
SELECT ST_Perimeter(ST_GeomFromText('POLYGON((743238 2967416,743238 2967450,743265 2967450, 743265.625 2967416,743238 2967416))', 2249)); st_perimeter --------- 122.630744000095 (1 row) SELECT ST_Perimeter(ST_GeomFromText('MULTIPOLYGON(((763104.471273676 2949418.44119003, 763104.477769673 2949418.42538203, 763104.189609677 2949418.22343004,763104.471273676 2949418.44119003)), ((763104.471273676 2949418.44119003,763095.804579742 2949436.33850239, 763086.132105649 2949451.46730207,763078.452329651 2949462.11549407, 763075.354136904 2949466.17407812,763064.362142565 2949477.64291974, 763059.953961626 2949481.28983009,762994.637609571 2949532.04103014, 762990.568508415 2949535.06640477,762986.710889563 2949539.61421415, 763117.237897679 2949709.50493431,763235.236617789 2949617.95619822, 763287.718121842 2949562.20592617,763111.553321674 2949423.91664605, 763104.471273676 2949418.44119003)))', 2249)); st_perimeter --------- 845.227713366825 (1 row)
Return perimeter in meters and feet for Polygon and MultiPolygon. Note this is geography (WGS 84 long lat)
SELECT ST_Perimeter(geog) As per_meters, ST_Perimeter(geog)/0.3048 As per_ft FROM ST_GeogFromText('POLYGON((-71.1776848522251 42.3902896512902,-71.1776843766326 42.3903829478009, -71.1775844305465 42.3903826677917,-71.1775825927231 42.3902893647987,-71.1776848522251 42.3902896512902))') As geog; per_meters | per_ft -----------------+------------------ 37.3790462565251 | 122.634666195949 -- MultiPolygon example -- SELECT ST_Perimeter(geog) As per_meters, ST_Perimeter(geog,false) As per_sphere_meters, ST_Perimeter(geog)/0.3048 As per_ft FROM ST_GeogFromText('MULTIPOLYGON(((-71.1044543107478 42.340674480411,-71.1044542869917 42.3406744369506, -71.1044553562977 42.340673886454,-71.1044543107478 42.340674480411)), ((-71.1044543107478 42.340674480411,-71.1044860600303 42.3407237015564,-71.1045215770124 42.3407653385914, -71.1045498002983 42.3407946553165,-71.1045611902745 42.3408058316308,-71.1046016507427 42.340837442371, -71.104617893173 42.3408475056957,-71.1048586153981 42.3409875993595,-71.1048736143677 42.3409959528211, -71.1048878050242 42.3410084812078,-71.1044020965803 42.3414730072048, -71.1039672113619 42.3412202916693,-71.1037740497748 42.3410666421308, -71.1044280218456 42.3406894151355,-71.1044543107478 42.340674480411)))') As geog; per_meters | per_sphere_meters | per_ft ------------------+-------------------+------------------ 257.634283683311 | 257.412311446337 | 845.256836231335
ST_Perimeter2D — Returns the 2-dimensional perimeter of the geometry, if it is a polygon or multi-polygon. This is currently an alias for ST_Perimeter.
float ST_Perimeter2D(
geometry geomA)
;
ST_3DPerimeter — Returns the 3-dimensional perimeter of the geometry, if it is a polygon or multi-polygon.
float ST_3DPerimeter(
geometry geomA)
;
Returns the 3-dimensional perimeter of the geometry, if it is a polygon or multi-polygon. If the geometry is 2-dimensional, then the 2-dimensional perimeter is returned.
This function supports 3d and will not drop the z-index.
Changed: 2.0.0 In prior versions this used to be called ST_Perimeter3D
Perimeter of a slightly elevated polygon in the air in Massachusetts state plane feet
SELECT ST_3DPerimeter(the_geom), ST_Perimeter2d(the_geom), ST_Perimeter(the_geom) FROM (SELECT ST_GeomFromEWKT('SRID=2249;POLYGON((743238 2967416 2,743238 2967450 1, 743265.625 2967416 1,743238 2967416 2))') As the_geom) As foo; ST_3DPerimeter | st_perimeter2d | st_perimeter ------------------+------------------+------------------ 105.465793597674 | 105.432997272188 | 105.432997272188
ST_PointOnSurface — Returns a POINT
guaranteed to lie on the surface.
geometry ST_PointOnSurface(
geometry g1)
;
Returns a POINT
guaranteed to intersect a surface.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s3.2.14.2 // s3.2.18.2
This method implements the SQL/MM specification. SQL-MM 3: 8.1.5, 9.5.6. According to the specs, ST_PointOnSurface works for surface geometries (POLYGONs, MULTIPOLYGONS, CURVED POLYGONS). So PostGIS seems to be extending what the spec allows here. Most databases Oracle,DB II, ESRI SDE seem to only support this function for surfaces. SQL Server 2008 like PostGIS supports for all common geometries.
This function supports 3d and will not drop the z-index.
SELECT ST_AsText(ST_PointOnSurface('POINT(0 5)'::geometry)); st_astext ------------ POINT(0 5) (1 row) SELECT ST_AsText(ST_PointOnSurface('LINESTRING(0 5, 0 10)'::geometry)); st_astext ------------ POINT(0 5) (1 row) SELECT ST_AsText(ST_PointOnSurface('POLYGON((0 0, 0 5, 5 5, 5 0, 0 0))'::geometry)); st_astext ---------------- POINT(2.5 2.5) (1 row) SELECT ST_AsEWKT(ST_PointOnSurface(ST_GeomFromEWKT('LINESTRING(0 5 1, 0 0 1, 0 10 2)'))); st_asewkt ---------------- POINT(0 0 1) (1 row)
ST_Project — Returns a POINT
projected from a start point using a distance in meters and bearing (azimuth) in radians.
geography ST_Project(
geography g1, float distance, float azimuth)
;
Returns a POINT
projected along a geodesic from a start point using an azimuth (bearing) measured in radians and distance measured in meters. This is also called a direct geodesic problem.
The azimuth is sometimes called the heading or the bearing in navigation. It is measured relative to true north (azimuth zero). East is azimuth 90 (π/2), south is azimuth 180 (π), west is azimuth 270 (3π/2).
The distance is given in meters.
Disponibilità: 2.0
ST_Relate — Returns true if this Geometry is spatially related to anotherGeometry, by testing for intersections between the Interior, Boundary and Exterior of the two geometries as specified by the values in the intersectionMatrixPattern. If no intersectionMatrixPattern is passed in, then returns the maximum intersectionMatrixPattern that relates the 2 geometries.
boolean ST_Relate(
geometry geomA, geometry geomB, text intersectionMatrixPattern)
;
text ST_Relate(
geometry geomA, geometry geomB)
;
text ST_Relate(
geometry geomA, geometry geomB, integer BoundaryNodeRule)
;
Version 1: Takes geomA, geomB, intersectionMatrix and Returns 1 (TRUE) if this Geometry is spatially related to anotherGeometry, by testing for intersections between the Interior, Boundary and Exterior of the two geometries as specified by the values in the DE-9IM matrix pattern.
This is especially useful for testing compound checks of intersection, crosses, etc in one step.
Do not call with a GeometryCollection as an argument
This is the "allowable" version that returns a boolean, not an integer. This is defined in OGC spec |
This DOES NOT automagically include an index call. The reason for that is some relationships are anti e.g. Disjoint. If you are using a relationship pattern that requires intersection, then include the && index call. |
Version 2: Takes geomA and geomB and returns the Section 4.3.6, “Dimensionally Extended 9 Intersection Model (DE-9IM)”
Version 3: same as version 2, but allows to specify a boundary node rule (1:OGC/MOD2, 2:Endpoint, 3:MultivalentEndpoint, 4:MonovalentEndpoint)
Do not call with a GeometryCollection as an argument |
not in OGC spec, but implied. see s2.1.13.2
Eseguito dal modulo GEOS
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2 // s2.1.13.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.25
Enhanced: 2.0.0 - added support for specifying boundary node rule (requires GEOS >= 3.0).
--Find all compounds that intersect and not touch a poly (interior intersects) SELECT l.* , b.name As poly_name FROM polys As b INNER JOIN compounds As l ON (p.the_geom && b.the_geom AND ST_Relate(l.the_geom, b.the_geom,'T********')); SELECT ST_Relate(ST_GeometryFromText('POINT(1 2)'), ST_Buffer(ST_GeometryFromText('POINT(1 2)'),2)); st_relate ----------- 0FFFFF212 SELECT ST_Relate(ST_GeometryFromText('LINESTRING(1 2, 3 4)'), ST_GeometryFromText('LINESTRING(5 6, 7 8)')); st_relate ----------- FF1FF0102 SELECT ST_Relate(ST_GeometryFromText('POINT(1 2)'), ST_Buffer(ST_GeometryFromText('POINT(1 2)'),2), '0FFFFF212'); st_relate ----------- t SELECT ST_Relate(ST_GeometryFromText('POINT(1 2)'), ST_Buffer(ST_GeometryFromText('POINT(1 2)'),2), '*FF*FF212'); st_relate ----------- t
ST_RelateMatch — Returns true if intersectionMattrixPattern1 implies intersectionMatrixPattern2
boolean ST_RelateMatch(
text intersectionMatrix, text intersectionMatrixPattern)
;
Takes intersectionMatrix and intersectionMatrixPattern and Returns true if the intersectionMatrix satisfies the intersectionMatrixPattern. For more information refer to Section 4.3.6, “Dimensionally Extended 9 Intersection Model (DE-9IM)”.
Availability: 2.0.0 - requires GEOS >= 3.3.0.
SELECT ST_RelateMatch('101202FFF', 'TTTTTTFFF') ; -- result -- t --example of common intersection matrix patterns and example matrices -- comparing relationships of involving one invalid geometry and ( a line and polygon that intersect at interior and boundary) SELECT mat.name, pat.name, ST_RelateMatch(mat.val, pat.val) As satisfied FROM ( VALUES ('Equality', 'T1FF1FFF1'), ('Overlaps', 'T*T***T**'), ('Within', 'T*F**F***'), ('Disjoint', 'FF*FF****') As pat(name,val) CROSS JOIN ( VALUES ('Self intersections (invalid)', '111111111'), ('IE2_BI1_BB0_BE1_EI1_EE2', 'FF2101102'), ('IB1_IE1_BB0_BE0_EI2_EI1_EE2', 'F11F00212') ) As mat(name,val);
ST_ShortestLine — Returns the 2-dimensional shortest line between two geometries
geometry ST_ShortestLine(
geometry g1, geometry g2)
;
Returns the 2-dimensional shortest line between two geometries. The function will only return the first shortest line if more than one, that the function finds. If g1 and g2 intersects in just one point the function will return a line with both start and end in that intersection-point. If g1 and g2 are intersecting with more than one point the function will return a line with start and end in the same point but it can be any of the intersecting points. The line returned will always start in g1 and end in g2. The length of the line this function returns will always be the same as ST_Distance returns for g1 and g2.
Disponibilità: 1.5.0
SELECT ST_AsText( ST_ShortestLine('POINT(100 100)'::geometry, 'LINESTRING (20 80, 98 190, 110 180, 50 75 )'::geometry) ) As sline; sline ----------------- LINESTRING(100 100,73.0769230769231 115.384615384615)
|
SELECT ST_AsText( ST_ShortestLine( ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) ) ) As slinewkt; LINESTRING(140.752120669087 125.695053378061,121.111404660392 153.370607753949)
|
ST_Touches — Returns TRUE
if the geometries have at least one point in common, but their interiors do not intersect.
boolean ST_Touches(
geometry g1, geometry g2)
;
Returns TRUE
if the only points in common between g1
and g2
lie in the union of the boundaries of g1
and g2
. The ST_Touches
relation applies to all Area/Area, Line/Line, Line/Area, Point/Area and Point/Line pairs of relationships, but not to the Point/Point pair.
In mathematical terms, this predicate is expressed as:
The allowable DE-9IM Intersection Matrices for the two geometries are:
FT*******
F**T*****
F***T****
Do not call with a |
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. To avoid using an index, use |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2 // s2.1.13.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.28
The ST_Touches
predicate returns TRUE
in all the following illustrations.
SELECT ST_Touches('LINESTRING(0 0, 1 1, 0 2)'::geometry, 'POINT(1 1)'::geometry); st_touches ------------ f (1 row) SELECT ST_Touches('LINESTRING(0 0, 1 1, 0 2)'::geometry, 'POINT(0 2)'::geometry); st_touches ------------ t (1 row)
ST_Within — Returns true if the geometry A is completely inside geometry B
boolean ST_Within(
geometry A, geometry B)
;
Returns TRUE if geometry A is completely inside geometry B. For this function to make sense, the source geometries must both be of the same coordinate projection, having the same SRID. It is a given that if ST_Within(A,B) is true and ST_Within(B,A) is true, then the two geometries are considered spatially equal.
Eseguito dal modulo GEOS
Enhanced: 2.3.0 Enhancement to PIP short-circuit for geometry extended to support MultiPoints with few points. Prior versions only supported point in polygon.
Do not call with a |
Do not use this function with invalid geometries. You will get unexpected results. |
This function call will automatically include a bounding box comparison that will make use of any indexes that are available on the geometries. To avoid index use, use the function _ST_Within.
NOTE: this is the "allowable" version that returns a boolean, not an integer.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.2 // s2.1.13.3 - a.Relate(b, 'T*F**F***')
This method implements the SQL/MM specification. SQL-MM 3: 5.1.30
--a circle within a circle SELECT ST_Within(smallc,smallc) As smallinsmall, ST_Within(smallc, bigc) As smallinbig, ST_Within(bigc,smallc) As biginsmall, ST_Within(ST_Union(smallc, bigc), bigc) as unioninbig, ST_Within(bigc, ST_Union(smallc, bigc)) as biginunion, ST_Equals(bigc, ST_Union(smallc, bigc)) as bigisunion FROM ( SELECT ST_Buffer(ST_GeomFromText('POINT(50 50)'), 20) As smallc, ST_Buffer(ST_GeomFromText('POINT(50 50)'), 40) As bigc) As foo; --Result smallinsmall | smallinbig | biginsmall | unioninbig | biginunion | bigisunion --------------+------------+------------+------------+------------+------------ t | t | f | t | t | t (1 row)
Abstract
SFCGAL is a C++ wrapper library around CGAL that provides advanced 2D and 3D functions. For robustness, geometry coordinates have an exact rational number representation.
Installation instructions of the library can be found on SFCGAL home page http://www.sfcgal.org. To load the functions create extension postgis_sfcgal.
Some SFCGAL functions replace standard ones (ST_Intersects, ST_Intersection, ST_Difference, ST_Union, ST_Area and ST_Distance), to switch between standard functions and SFCGAL function use:
SET postgis.backend = sfcgal;
and
SET postgis.backend = geos;
postgis_sfcgal_version — Returns the version of SFCGAL in use
text postgis_sfcgal_version(
void)
;
ST_Extrude — Estrude una superficie a volume
geometry ST_Extrude(
geometry geom, float x, float y, float z)
;
Availability: 2.1.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
3D images were generated using PostGIS ST_AsX3D and rendering in HTML using X3Dom HTML Javascript rendering library.
SELECT ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30);
|
ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30);
|
SELECT ST_GeomFromText('LINESTRING(50 50, 100 90, 95 150)')
|
SELECT ST_Extrude( ST_GeomFromText('LINESTRING(50 50, 100 90, 95 150)'),0,0,10));
|
ST_StraightSkeleton — Compute a straight skeleton from a geometry
geometry ST_StraightSkeleton(
geometry geom)
;
ST_ApproximateMedialAxis — Compute the approximate medial axis of an areal geometry.
geometry ST_ApproximateMedialAxis(
geometry geom)
;
Return an approximate medial axis for the areal input based on its straight skeleton. Uses an SFCGAL specific API when built against a capable version (1.2.0+). Otherwise the function is just a wrapper around ST_StraightSkeleton (slower case).
Availability: 2.2.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_IsPlanar — Verifica se una superficie è planare o meno.
boolean ST_IsPlanar(
geometry geom)
;
Availability: 2.2.0: This was documented in 2.1.0 but got accidentally left out in 2.1 release.
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_Orientation — Determina l'orientazione di una superficie.
integer ST_Orientation(
geometry geom)
;
ST_MinkowskiSum — Performs Minkowski sum
geometry ST_MinkowskiSum(
geometry geom1, geometry geom2)
;
This function performs a 2D minkowski sum of a point, line or polygon with a polygon.
A minkowski sum of two geometries A and B is the set of all points that are the sum of any point in A and B. Minkowski sums are often used in motion planning and computer-aided design. More details on Wikipedia Minkowski addition.
The first parameter can be any 2D geometry (point, linestring, polygon). If a 3D geometry is passed, it will be converted to 2D by forcing Z to 0, leading to possible cases of invalidity. The second parameter must be a 2D polygon.
Implementation utilizes CGAL 2D Minkowskisum.
Availability: 2.1.0
This method needs SFCGAL backend.
Minkowski Sum of Linestring and circle polygon where Linestring cuts thru the circle
|
|
SELECT ST_MinkowskiSum(line, circle)) FROM (SELECT ST_MakeLine(ST_MakePoint(10, 10),ST_MakePoint(100, 100)) As line, ST_Buffer(ST_GeomFromText('POINT(50 50)'), 30) As circle) As foo; -- wkt -- MULTIPOLYGON(((30 59.9999999999999,30.5764415879031 54.1472903395161,32.2836140246614 48.5194970290472,35.0559116309237 43.3328930094119,38.7867965644036 38.7867965644035,43.332893009412 35.0559116309236,48.5194970290474 32.2836140246614,54.1472903395162 30.5764415879031,60.0000000000001 30,65.8527096604839 30.5764415879031,71.4805029709527 32.2836140246614,76.6671069905881 35.0559116309237,81.2132034355964 38.7867965644036,171.213203435596 128.786796564404,174.944088369076 133.332893009412,177.716385975339 138.519497029047,179.423558412097 144.147290339516,180 150,179.423558412097 155.852709660484,177.716385975339 161.480502970953,174.944088369076 166.667106990588,171.213203435596 171.213203435596,166.667106990588 174.944088369076, 161.480502970953 177.716385975339,155.852709660484 179.423558412097,150 180,144.147290339516 179.423558412097,138.519497029047 177.716385975339,133.332893009412 174.944088369076,128.786796564403 171.213203435596,38.7867965644035 81.2132034355963,35.0559116309236 76.667106990588,32.2836140246614 71.4805029709526,30.5764415879031 65.8527096604838,30 59.9999999999999)))
Minkowski Sum of a polygon and multipoint
|
|
SELECT ST_MinkowskiSum(mp, poly) FROM (SELECT 'MULTIPOINT(25 50,70 25)'::geometry As mp, 'POLYGON((130 150, 20 40, 50 60, 125 100, 130 150))'::geometry As poly ) As foo -- wkt -- MULTIPOLYGON( ((70 115,100 135,175 175,225 225,70 115)), ((120 65,150 85,225 125,275 175,120 65)) )
ST_3DIntersection — Perform 3D intersection
geometry ST_3DIntersection(
geometry geom1, geometry geom2)
;
Return a geometry that is the shared portion between geom1 and geom2.
Availability: 2.1.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
3D images were generated using PostGIS ST_AsX3D and rendering in HTML using X3Dom HTML Javascript rendering library.
SELECT ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30) AS geom1, ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(80 80)'), 50, 'quad_segs=1'),0,0,30) AS geom2;
|
SELECT ST_3DIntersection(geom1,geom2) FROM ( SELECT ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30) AS geom1, ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(80 80)'), 50, 'quad_segs=1'),0,0,30) AS geom2 ) As t;
|
3D linestrings and polygons
SELECT ST_AsText(ST_3DIntersection(linestring, polygon)) As wkt FROM ST_GeomFromText('LINESTRING Z (2 2 6,1.5 1.5 7,1 1 8,0.5 0.5 8,0 0 10)') AS linestring CROSS JOIN ST_GeomFromText('POLYGON((0 0 8, 0 1 8, 1 1 8, 1 0 8, 0 0 8))') AS polygon; wkt -------------------------------- LINESTRING Z (1 1 8,0.5 0.5 8)
Cube (closed Polyhedral Surface) and Polygon Z
SELECT ST_AsText(ST_3DIntersection( ST_GeomFromText('POLYHEDRALSURFACE Z( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )'), 'POLYGON Z ((0 0 0, 0 0 0.5, 0 0.5 0.5, 0 0.5 0, 0 0 0))'::geometry))
TIN Z (((0 0 0,0 0 0.5,0 0.5 0.5,0 0 0)),((0 0.5 0,0 0 0,0 0.5 0.5,0 0.5 0)))
Intersection of 2 solids that result in volumetric intersection is also a solid (ST_Dimension returns 3)
SELECT ST_AsText(ST_3DIntersection( ST_Extrude(ST_Buffer('POINT(10 20)'::geometry,10,1),0,0,30), ST_Extrude(ST_Buffer('POINT(10 20)'::geometry,10,1),2,0,10) ));
POLYHEDRALSURFACE Z (((13.3333333333333 13.3333333333333 10,20 20 0,20 20 10,13.3333333333333 13.3333333333333 10)), ((20 20 10,16.6666666666667 23.3333333333333 10,13.3333333333333 13.3333333333333 10,20 20 10)), ((20 20 0,16.6666666666667 23.3333333333333 10,20 20 10,20 20 0)), ((13.3333333333333 13.3333333333333 10,10 10 0,20 20 0,13.3333333333333 13.3333333333333 10)), ((16.6666666666667 23.3333333333333 10,12 28 10,13.3333333333333 13.3333333333333 10,16.6666666666667 23.3333333333333 10)), ((20 20 0,9.99999999999995 30 0,16.6666666666667 23.3333333333333 10,20 20 0)), ((10 10 0,9.99999999999995 30 0,20 20 0,10 10 0)),((13.3333333333333 13.3333333333333 10,12 12 10,10 10 0,13.3333333333333 13.3333333333333 10)), ((12 28 10,12 12 10,13.3333333333333 13.3333333333333 10,12 28 10)), ((16.6666666666667 23.3333333333333 10,9.99999999999995 30 0,12 28 10,16.6666666666667 23.3333333333333 10)), ((10 10 0,0 20 0,9.99999999999995 30 0,10 10 0)), ((12 12 10,11 11 10,10 10 0,12 12 10)),((12 28 10,11 11 10,12 12 10,12 28 10)), ((9.99999999999995 30 0,11 29 10,12 28 10,9.99999999999995 30 0)),((0 20 0,2 20 10,9.99999999999995 30 0,0 20 0)), ((10 10 0,2 20 10,0 20 0,10 10 0)),((11 11 10,2 20 10,10 10 0,11 11 10)),((12 28 10,11 29 10,11 11 10,12 28 10)), ((9.99999999999995 30 0,2 20 10,11 29 10,9.99999999999995 30 0)),((11 11 10,11 29 10,2 20 10,11 11 10)))
ST_3DDifference — Perform 3D difference
geometry ST_3DDifference(
geometry geom1, geometry geom2)
;
Returns that part of geom1 that is not part of geom2.
Availability: 2.2.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
3D images were generated using PostGIS ST_AsX3D and rendering in HTML using X3Dom HTML Javascript rendering library.
SELECT ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30) AS geom1, ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(80 80)'), 50, 'quad_segs=1'),0,0,30) AS geom2;
|
SELECT ST_3DDifference(geom1,geom2) FROM ( SELECT ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30) AS geom1, ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(80 80)'), 50, 'quad_segs=1'),0,0,30) AS geom2 ) As t;
|
ST_3DUnion — Perform 3D union
geometry ST_3DUnion(
geometry geom1, geometry geom2)
;
Availability: 2.2.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
3D images were generated using PostGIS ST_AsX3D and rendering in HTML using X3Dom HTML Javascript rendering library.
SELECT ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30) AS geom1, ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(80 80)'), 50, 'quad_segs=1'),0,0,30) AS geom2;
|
SELECT ST_3DUnion(geom1,geom2) FROM ( SELECT ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2'),0,0,30) AS geom1, ST_Extrude(ST_Buffer(ST_GeomFromText('POINT(80 80)'), 50, 'quad_segs=1'),0,0,30) AS geom2 ) As t;
|
ST_3DArea — Computes area of 3D surface geometries. Will return 0 for solids.
floatST_3DArea(
geometry geom1)
;
Availability: 2.1.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
Note: By default a PolyhedralSurface built from WKT is a surface geometry, not solid. It therefore has surface area. Once converted to a solid, no area.
SELECT ST_3DArea(geom) As cube_surface_area, ST_3DArea(ST_MakeSolid(geom)) As solid_surface_area FROM (SELECT 'POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )'::geometry) As f(geom); cube_surface_area | solid_surface_area -------------------+-------------------- 6 | 0
ST_Tesselate — Perform surface Tesselation of a polygon or polyhedralsurface and returns as a TIN or collection of TINS
geometry ST_Tesselate(
geometry geom)
;
Takes as input a surface such a MULTI(POLYGON) or POLYHEDRALSURFACE and returns a TIN representation via the process of tesselation using triangles.
Availability: 2.1.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT ST_GeomFromText('POLYHEDRALSURFACE Z( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )');
|
SELECT ST_Tesselate(ST_GeomFromText('POLYHEDRALSURFACE Z( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )')); ST_AsText output: TIN Z (((0 0 0,0 0 1,0 1 1,0 0 0)),((0 1 0,0 0 0,0 1 1,0 1 0)), ((0 0 0,0 1 0,1 1 0,0 0 0)), ((1 0 0,0 0 0,1 1 0,1 0 0)),((0 0 1,1 0 0,1 0 1,0 0 1)), ((0 0 1,0 0 0,1 0 0,0 0 1)), ((1 1 0,1 1 1,1 0 1,1 1 0)),((1 0 0,1 1 0,1 0 1,1 0 0)), ((0 1 0,0 1 1,1 1 1,0 1 0)),((1 1 0,0 1 0,1 1 1,1 1 0)), ((0 1 1,1 0 1,1 1 1,0 1 1)),((0 1 1,0 0 1,1 0 1,0 1 1)))
|
SELECT 'POLYGON (( 10 190, 10 70, 80 70, 80 130, 50 160, 120 160, 120 190, 10 190 ))'::geometry;
|
SELECT ST_Tesselate('POLYGON (( 10 190, 10 70, 80 70, 80 130, 50 160, 120 160, 120 190, 10 190 ))'::geometry);
ST_AsText output TIN(((80 130,50 160,80 70,80 130)),((50 160,10 190,10 70,50 160)), ((80 70,50 160,10 70,80 70)),((120 160,120 190,50 160,120 160)), ((120 190,10 190,50 160,120 190))) |
ST_Volume — Computes the volume of a 3D solid. If applied to surface (even closed) geometries will return 0.
float ST_Volume(
geometry geom1)
;
Availability: 2.2.0
This method needs SFCGAL backend.
This function supports 3d and will not drop the z-index.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
When closed surfaces are created with WKT, they are treated as areal rather than solid. To make them solid, you need to use ST_MakeSolid. Areal geometries have no volume. Here is an example to demonstrate.
SELECT ST_Volume(geom) As cube_surface_vol, ST_Volume(ST_MakeSolid(geom)) As solid_surface_vol FROM (SELECT 'POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )'::geometry) As f(geom); cube_surface_vol | solid_surface_vol ------------------+------------------- 0 | 1
ST_MakeSolid — Cast the geometry into a solid. No check is performed. To obtain a valid solid, the input geometry must be a closed Polyhedral Surface or a closed TIN.
geometryST_MakeSolid(
geometry geom1)
;
geometry_dump
rows, representing the exterior and interior rings of a polygon.ST_Buffer — (T) Returns a geometry covering all points within a given distance from the input geometry.
geometry ST_Buffer(
geometry g1, float radius_of_buffer)
;
geometry ST_Buffer(
geometry g1, float radius_of_buffer, integer num_seg_quarter_circle)
;
geometry ST_Buffer(
geometry g1, float radius_of_buffer, text buffer_style_parameters)
;
geography ST_Buffer(
geography g1, float radius_of_buffer_in_meters)
;
Returns a geometry/geography that represents all points whose distance from this Geometry/geography is less than or equal to distance.
Geometry: Calculations are in the Spatial Reference System of the geometry. Introduced in 1.5 support for different end cap and mitre settings to control shape.
Negative radii: For polygons, a negative radius can be used, which will shrink the polygon rather than expanding it. |
Geography: For geography this is really a thin wrapper around the geometry implementation. It first determines the best SRID that fits the bounding box of the geography object (favoring UTM, Lambert Azimuthal Equal Area (LAEA) north/south pole, and falling back on mercator in worst case scenario) and then buffers in that planar spatial ref and retransforms back to WGS84 geography. |
For geography this may not behave as expected if object is sufficiently large that it falls between two UTM zones or crosses the dateline
Availability: 1.5 - ST_Buffer was enhanced to support different endcaps and join types. These are useful for example to convert road linestrings into polygon roads with flat or square edges instead of rounded edges. Thin wrapper for geography was added. - requires GEOS >= 3.2 to take advantage of advanced geometry functionality.
The optional third parameter (currently only applies to geometry) can either specify number of segments used to approximate a quarter circle (integer case, defaults to 8) or a list of blank-separated key=value pairs (string case) to tweak operations as follows:
'quad_segs=#' : number of segments used to approximate a quarter circle (defaults to 8).
'endcap=round|flat|square' : endcap style (defaults to "round", needs GEOS-3.2 or higher for a different value). 'butt' is also accepted as a synonym for 'flat'.
'join=round|mitre|bevel' : join style (defaults to "round", needs GEOS-3.2 or higher for a different value). 'miter' is also accepted as a synonym for 'mitre'.
'mitre_limit=#.#' : mitre ratio limit (only affects mitered join style). 'miter_limit' is also accepted as a synonym for 'mitre_limit'.
Units of radius are measured in units of the spatial reference system.
The inputs can be POINTS, MULTIPOINTS, LINESTRINGS, MULTILINESTRINGS, POLYGONS, MULTIPOLYGONS, and GeometryCollections.
This function ignores the third dimension (z) and will always give a 2-d buffer even when presented with a 3d-geometry. |
Eseguito dal modulo GEOS
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.17
People often make the mistake of using this function to try to do radius searches. Creating a buffer to to a radius search is slow and pointless. Use ST_DWithin instead. |
SELECT ST_Buffer( ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=8');
|
SELECT ST_Buffer( ST_GeomFromText('POINT(100 90)'), 50, 'quad_segs=2');
| |
SELECT ST_Buffer( ST_GeomFromText( 'LINESTRING(50 50,150 150,150 50)' ), 10, 'endcap=round join=round');
|
SELECT ST_Buffer( ST_GeomFromText( 'LINESTRING(50 50,150 150,150 50)' ), 10, 'endcap=square join=round');
|
SELECT ST_Buffer( ST_GeomFromText( 'LINESTRING(50 50,150 150,150 50)' ), 10, 'endcap=flat join=round');
|
SELECT ST_Buffer( ST_GeomFromText( 'LINESTRING(50 50,150 150,150 50)' ), 10, 'join=bevel');
|
SELECT ST_Buffer( ST_GeomFromText( 'LINESTRING(50 50,150 150,150 50)' ), 10, 'join=mitre mitre_limit=5.0');
|
SELECT ST_Buffer( ST_GeomFromText( 'LINESTRING(50 50,150 150,150 50)' ), 10, 'join=mitre mitre_limit=1.0');
|
--A buffered point approximates a circle -- A buffered point forcing approximation of (see diagram) -- 2 points per circle is poly with 8 sides (see diagram) SELECT ST_NPoints(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50)) As promisingcircle_pcount, ST_NPoints(ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50, 2)) As lamecircle_pcount; promisingcircle_pcount | lamecircle_pcount ------------------------+------------------- 33 | 9 --A lighter but lamer circle -- only 2 points per quarter circle is an octagon --Below is a 100 meter octagon -- Note coordinates are in NAD 83 long lat which we transform to Mass state plane meter and then buffer to get measurements in meters; SELECT ST_AsText(ST_Buffer( ST_Transform( ST_SetSRID(ST_MakePoint(-71.063526, 42.35785),4269), 26986) ,100,2)) As octagon; ---------------------- POLYGON((236057.59057465 900908.759918696,236028.301252769 900838.049240578,235 957.59057465 900808.759918696,235886.879896532 900838.049240578,235857.59057465 900908.759918696,235886.879896532 900979.470596815,235957.59057465 901008.759918 696,236028.301252769 900979.470596815,236057.59057465 900908.759918696))
ST_BuildArea — Creates an areal geometry formed by the constituent linework of given geometry
geometry ST_BuildArea(
geometry A)
;
Creates an areal geometry formed by the constituent linework of given geometry. The return type can be a Polygon or MultiPolygon, depending on input. If the input lineworks do not form polygons NULL is returned. The inputs can be LINESTRINGS, MULTILINESTRINGS, POLYGONS, MULTIPOLYGONS, and GeometryCollections.
This function will assume all inner geometries represent holes
Input linework must be correctly noded for this function to work properly |
Availability: 1.1.0 - requires GEOS >= 2.1.0.
SELECT ST_BuildArea(ST_Collect(smallc,bigc)) FROM (SELECT ST_Buffer( ST_GeomFromText('POINT(100 90)'), 25) As smallc, ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50) As bigc) As foo;
|
SELECT ST_BuildArea(ST_Collect(line,circle)) FROM (SELECT ST_Buffer( ST_MakeLine(ST_MakePoint(10, 10),ST_MakePoint(190, 190)), 5) As line, ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50) As circle) As foo; --this creates the same gaping hole --but using linestrings instead of polygons SELECT ST_BuildArea( ST_Collect(ST_ExteriorRing(line),ST_ExteriorRing(circle)) ) FROM (SELECT ST_Buffer( ST_MakeLine(ST_MakePoint(10, 10),ST_MakePoint(190, 190)) ,5) As line, ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50) As circle) As foo;
|
ST_Node, ST_MakePolygon, ST_BdPolyFromText, ST_BdMPolyFromTextwrappers to this function with standard OGC interface
ST_ClipByBox2D — Returns the portion of a geometry falling within a rectangle.
geometry ST_ClipByBox2D(
geometry geom, box2d box)
;
Clips a geometry by a 2D box in a fast but possibly dirty way. The output geometry is not guaranteed to be valid (self-intersections for a polygon may be introduced). Topologically invalid input geometries do not result in exceptions being thrown.
Eseguito dal modulo GEOS
Requires GEOS 3.5.0+ |
Availability: 2.2.0 - requires GEOS >= 3.5.0.
ST_ClusterIntersecting — Aggregate. Returns an array with the connected components of a set of geometries
geometry[] ST_ClusterIntersecting(
geometry g)
;
ST_ClusterIntersecting is an aggregate function that returns an array of GeometryCollections, where each GeometryCollection represents an interconnected set of geometries.
Availability: 2.2.0 - requires GEOS
WITH testdata AS (SELECT unnest(ARRAY['LINESTRING (0 0, 1 1)'::geometry, 'LINESTRING (5 5, 4 4)'::geometry, 'LINESTRING (6 6, 7 7)'::geometry, 'LINESTRING (0 0, -1 -1)'::geometry, 'POLYGON ((0 0, 4 0, 4 4, 0 4, 0 0))'::geometry]) AS geom) SELECT ST_AsText(unnest(ST_ClusterIntersecting(geom))) FROM testdata; --result st_astext --------- GEOMETRYCOLLECTION(LINESTRING(0 0,1 1),LINESTRING(5 5,4 4),LINESTRING(0 0,-1 -1),POLYGON((0 0,4 0,4 4,0 4,0 0))) GEOMETRYCOLLECTION(LINESTRING(6 6,7 7))
ST_ClusterWithin — Aggregate. Returns an array of GeometryCollections, where each GeometryCollection represents a set of geometries separated by no more than the specified distance.
geometry[] ST_ClusterWithin(
geometry g, float8 distance)
;
ST_ClusterWithin is an aggregate function that returns an array of GeometryCollections, where each GeometryCollection represents a set of geometries separated by no more than the specified distance.
Availability: 2.2.0 - requires GEOS
WITH testdata AS (SELECT unnest(ARRAY['LINESTRING (0 0, 1 1)'::geometry, 'LINESTRING (5 5, 4 4)'::geometry, 'LINESTRING (6 6, 7 7)'::geometry, 'LINESTRING (0 0, -1 -1)'::geometry, 'POLYGON ((0 0, 4 0, 4 4, 0 4, 0 0))'::geometry]) AS geom) SELECT ST_AsText(unnest(ST_ClusterWithin(geom, 1.4))) FROM testdata; --result st_astext --------- GEOMETRYCOLLECTION(LINESTRING(0 0,1 1),LINESTRING(5 5,4 4),LINESTRING(0 0,-1 -1),POLYGON((0 0,4 0,4 4,0 4,0 0))) GEOMETRYCOLLECTION(LINESTRING(6 6,7 7))
ST_Collect — Return a specified ST_Geometry value from a collection of other geometries.
geometry ST_Collect(
geometry set g1field)
;
geometry ST_Collect(
geometry g1, geometry g2)
;
geometry ST_Collect(
geometry[] g1_array)
;
Output type can be a MULTI* or a GEOMETRYCOLLECTION. Comes in 2 variants. Variant 1 collects 2 geometries. Variant 2 is an aggregate function that takes a set of geometries and collects them into a single ST_Geometry.
Aggregate version: This function returns a GEOMETRYCOLLECTION or a MULTI object from a set of geometries. The ST_Collect() function is an "aggregate" function in the terminology of PostgreSQL. That means that it operates on rows of data, in the same way the SUM() and AVG() functions do. For example, "SELECT ST_Collect(GEOM) FROM GEOMTABLE GROUP BY ATTRCOLUMN" will return a separate GEOMETRYCOLLECTION for each distinct value of ATTRCOLUMN.
Non-Aggregate version: This function returns a geometry being a collection of two input geometries. Output type can be a MULTI* or a GEOMETRYCOLLECTION.
ST_Collect and ST_Union are often interchangeable. ST_Collect is in general orders of magnitude faster than ST_Union because it does not try to dissolve boundaries or validate that a constructed MultiPolgon doesn't have overlapping regions. It merely rolls up single geometries into MULTI and MULTI or mixed geometry types into Geometry Collections. Unfortunately geometry collections are not well-supported by GIS tools. To prevent ST_Collect from returning a Geometry Collection when collecting MULTI geometries, one can use the below trick that utilizes ST_Dump to expand the MULTIs out to singles and then regroup them. |
Availability: 1.4.0 - ST_Collect(geomarray) was introduced. ST_Collect was enhanced to handle more geometries faster.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves This method supports Circular Strings and Curves, but will never return a MULTICURVE or MULTI as one would expect and PostGIS does not currently support those.
Aggregate example (http://postgis.refractions.net/pipermail/postgis-users/2008-June/020331.html)
SELECT stusps, ST_Multi(ST_Collect(f.the_geom)) as singlegeom FROM (SELECT stusps, (ST_Dump(the_geom)).geom As the_geom FROM somestatetable ) As f GROUP BY stusps
Non-Aggregate example
SELECT ST_AsText(ST_Collect(ST_GeomFromText('POINT(1 2)'), ST_GeomFromText('POINT(-2 3)') )); st_astext ---------- MULTIPOINT(1 2,-2 3) --Recupera punti 2D SELECT ST_AsText(ST_Collect(ST_GeomFromText('POINT(1 2)'), ST_GeomFromText('POINT(1 2)') ) ); st_astext ---------- MULTIPOINT(1 2,1 2) --recupera punti 3D SELECT ST_AsEWKT(ST_Collect(ST_GeomFromEWKT('POINT(1 2 3)'), ST_GeomFromEWKT('POINT(1 2 4)') ) ); st_asewkt ------------------------- MULTIPOINT(1 2 3,1 2 4) --esempio con curve SELECT ST_AsText(ST_Collect(ST_GeomFromText('CIRCULARSTRING(220268 150415,220227 150505,220227 150406)'), ST_GeomFromText('CIRCULARSTRING(220227 150406,2220227 150407,220227 150406)'))); st_astext ------------------------------------------------------------------------------------ GEOMETRYCOLLECTION(CIRCULARSTRING(220268 150415,220227 150505,220227 150406), CIRCULARSTRING(220227 150406,2220227 150407,220227 150406)) --nuova costruzione di ST_Collect array SELECT ST_Collect(ARRAY(SELECT the_geom FROM sometable)); SELECT ST_AsText(ST_Collect(ARRAY[ST_GeomFromText('LINESTRING(1 2, 3 4)'), ST_GeomFromText('LINESTRING(3 4, 4 5)')])) As wktcollect; --wkt collect -- MULTILINESTRING((1 2,3 4),(3 4,4 5))
ST_ConcaveHull — The concave hull of a geometry represents a possibly concave geometry that encloses all geometries within the set. You can think of it as shrink wrapping.
geometry ST_ConcaveHull(
geometry geomA, float target_percent, boolean allow_holes=false)
;
The concave hull of a geometry represents a possibly concave geometry that encloses all geometries within the set. Defaults to false for allowing polygons with holes. The result is never higher than a single polygon.
The target_percent is the target percent of area of convex hull the PostGIS solution will try to approach before giving up or exiting. One can think of the concave hull as the geometry you get by vacuum sealing a set of geometries. The target_percent of 1 will give you the same answer as the convex hull. A target_percent between 0 and 0.99 will give you something that should have a smaller area than the convex hull. This is different from a convex hull which is more like wrapping a rubber band around the set of geometries.
It is usually used with MULTI and Geometry Collections. Although it is not an aggregate - you can use it in conjunction with ST_Collect or ST_Union to get the concave hull of a set of points/linestring/polygons ST_ConcaveHull(ST_Collect(somepointfield), 0.80).
It is much slower to compute than convex hull but encloses the geometry better and is also useful for image recognition.
Eseguito dal modulo GEOS
Note - If you are using with points, linestrings, or geometry collections use ST_Collect. If you are using with polygons, use ST_Union since it may fail with invalid geometries. |
Note - The smaller you make the target percent, the longer it takes to process the concave hull and more likely to run into topological exceptions. Also the more floating points and number of points you accrue. First try a 0.99 which does a first hop, is usually very fast, sometimes as fast as computing the convex hull, and usually gives much better than 99% of shrink since it almost always overshoots. Second hope of 0.98 it slower, others get slower usually quadratically. To reduce precision and float points, use ST_SimplifyPreserveTopology or ST_SnapToGrid after ST_ConcaveHull. ST_SnapToGrid is a bit faster, but could result in invalid geometries where as ST_SimplifyPreserveTopology almost always preserves the validity of the geometry. |
More real world examples and brief explanation of the technique are shown http://www.bostongis.com/postgis_concavehull.snippet
Also check out Simon Greener's article on demonstrating ConcaveHull introduced in Oracle 11G R2. http://www.spatialdbadvisor.com/oracle_spatial_tips_tricks/172/concave-hull-geometries-in-oracle-11gr2. The solution we get at 0.75 target percent of convex hull is similar to the shape Simon gets with Oracle SDO_CONCAVEHULL_BOUNDARY.
Disponibilità: 2.0
--Get estimate of infected area based on point observations SELECT d.disease_type, ST_ConcaveHull(ST_Collect(d.pnt_geom), 0.99) As geom FROM disease_obs As d GROUP BY d.disease_type;
-- geometries overlaid with concavehull -- at target 100% shrink (this is the same as convex hull - since no shrink) SELECT ST_ConcaveHull( ST_Union(ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) ), 1) As convexhull;
|
-- geometries overlaid with concavehull at target 90% shrink SELECT ST_ConcaveHull( ST_Union(ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) ), 0.9) As target_90;
|
-- this produces a table of 42 points that form an L shape SELECT (ST_DumpPoints(ST_GeomFromText( 'MULTIPOINT(14 14,34 14,54 14,74 14,94 14,114 14,134 14, 150 14,154 14,154 6,134 6,114 6,94 6,74 6,54 6,34 6, 14 6,10 6,8 6,7 7,6 8,6 10,6 30,6 50,6 70,6 90,6 110,6 130, 6 150,6 170,6 190,6 194,14 194,14 174,14 154,14 134,14 114, 14 94,14 74,14 54,14 34,14 14)'))).geom INTO TABLE l_shape; SELECT ST_ConvexHull(ST_Collect(geom)) FROM l_shape;
|
SELECT ST_ConcaveHull(ST_Collect(geom), 0.99) FROM l_shape;
|
-- Concave Hull L shape points -- at target 80% of convexhull SELECT ST_ConcaveHull(ST_Collect(geom), 0.80) FROM l_shape;
|
SELECT ST_ConcaveHull(ST_GeomFromText('MULTILINESTRING((106 164,30 112,74 70,82 112,130 94, 130 62,122 40,156 32,162 76,172 88), (132 178,134 148,128 136,96 128,132 108,150 130, 170 142,174 110,156 96,158 90,158 88), (22 64,66 28,94 38,94 68,114 76,112 30, 132 10,168 18,178 34,186 52,184 74,190 100, 190 122,182 148,178 170,176 184,156 164,146 178, 132 186,92 182,56 158,36 150,62 150,76 128,88 118))'),0.99)
|
ST_ConvexHull — The convex hull of a geometry represents the minimum convex geometry that encloses all geometries within the set.
geometry ST_ConvexHull(
geometry geomA)
;
The convex hull of a geometry represents the minimum convex geometry that encloses all geometries within the set.
One can think of the convex hull as the geometry you get by wrapping an elastic band around a set of geometries. This is different from a concave hull which is analogous to shrink-wrapping your geometries.
It is usually used with MULTI and Geometry Collections. Although it is not an aggregate - you can use it in conjunction with ST_Collect to get the convex hull of a set of points. ST_ConvexHull(ST_Collect(somepointfield)).
It is often used to determine an affected area based on a set of point observations.
Eseguito dal modulo GEOS
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.16
This function supports 3d and will not drop the z-index.
--Get estimate of infected area based on point observations SELECT d.disease_type, ST_ConvexHull(ST_Collect(d.the_geom)) As the_geom FROM disease_obs As d GROUP BY d.disease_type;
SELECT ST_AsText(ST_ConvexHull( ST_Collect( ST_GeomFromText('MULTILINESTRING((100 190,10 8),(150 10, 20 30))'), ST_GeomFromText('MULTIPOINT(50 5, 150 30, 50 10, 10 10)') )) ); ---st_astext-- POLYGON((50 5,10 8,10 10,100 190,150 30,150 10,50 5))
ST_CurveToLine — Converte un CIRCULARSTRING/CURVEDPOLYGON in LINESTRING/POLYGON
geometry ST_CurveToLine(
geometry curveGeom)
;
geometry ST_CurveToLine(
geometry curveGeom, integer segments_per_qtr_circle)
;
Converst a CIRCULAR STRING to regular LINESTRING or CURVEPOLYGON to POLYGON. Useful for outputting to devices that can't support CIRCULARSTRING geometry types
Converts a given geometry to a linear geometry. Each curved geometry or segment is converted into a linear approximation using the default value of 32 segments per quarter circle
Availability: 1.2.2?
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1.
This method implements the SQL/MM specification. SQL-MM 3: 7.1.7
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_AsText(ST_CurveToLine(ST_GeomFromText('CIRCULARSTRING(220268 150415,220227 150505,220227 150406)'))); --Result -- LINESTRING(220268 150415,220269.95064912 150416.539364228,220271.823415575 150418.17258804,220273.613787707 150419.895736857, 220275.317452352 150421.704659462,220276.930305234 150423.594998003,220278.448460847 150425.562198489, 220279.868261823 150427.60152176,220281.186287736 150429.708054909,220282.399363347 150431.876723113, 220283.50456625 150434.10230186,220284.499233914 150436.379429536,220285.380970099 150438.702620341,220286.147650624 150441.066277505, 220286.797428488 150443.464706771,220287.328738321 150445.892130112,220287.740300149 150448.342699654, 220288.031122486 150450.810511759,220288.200504713 150453.289621251,220288.248038775 150455.77405574, 220288.173610157 150458.257830005,220287.977398166 150460.734960415,220287.659875492 150463.199479347, 220287.221807076 150465.64544956,220286.664248262 150468.066978495,220285.988542259 150470.458232479,220285.196316903 150472.81345077, 220284.289480732 150475.126959442,220283.270218395 150477.39318505,220282.140985384 150479.606668057, 220280.90450212 150481.762075989,220279.5637474 150483.85421628,220278.12195122 150485.87804878, 220276.582586992 150487.828697901,220274.949363179 150489.701464356,220273.226214362 150491.491836488, 220271.417291757 150493.195501133,220269.526953216 150494.808354014,220267.559752731 150496.326509628, 220265.520429459 150497.746310603,220263.41389631 150499.064336517,220261.245228106 150500.277412127, 220259.019649359 150501.38261503,220256.742521683 150502.377282695,220254.419330878 150503.259018879, 220252.055673714 150504.025699404,220249.657244448 150504.675477269,220247.229821107 150505.206787101, 220244.779251566 150505.61834893,220242.311439461 150505.909171266,220239.832329968 150506.078553494, 220237.347895479 150506.126087555,220234.864121215 150506.051658938,220232.386990804 150505.855446946, 220229.922471872 150505.537924272,220227.47650166 150505.099855856,220225.054972724 150504.542297043, 220222.663718741 150503.86659104,220220.308500449 150503.074365683, 220217.994991777 150502.167529512,220215.72876617 150501.148267175, 220213.515283163 150500.019034164,220211.35987523 150498.7825509, 220209.267734939 150497.441796181,220207.243902439 150496, 220205.293253319 150494.460635772,220203.420486864 150492.82741196,220201.630114732 150491.104263143, 220199.926450087 150489.295340538,220198.313597205 150487.405001997,220196.795441592 150485.437801511, 220195.375640616 150483.39847824,220194.057614703 150481.291945091,220192.844539092 150479.123276887,220191.739336189 150476.89769814, 220190.744668525 150474.620570464,220189.86293234 150472.297379659,220189.096251815 150469.933722495, 220188.446473951 150467.535293229,220187.915164118 150465.107869888,220187.50360229 150462.657300346, 220187.212779953 150460.189488241,220187.043397726 150457.710378749,220186.995863664 150455.22594426, 220187.070292282 150452.742169995,220187.266504273 150450.265039585,220187.584026947 150447.800520653, 220188.022095363 150445.35455044,220188.579654177 150442.933021505,220189.25536018 150440.541767521, 220190.047585536 150438.18654923,220190.954421707 150435.873040558,220191.973684044 150433.60681495, 220193.102917055 150431.393331943,220194.339400319 150429.237924011,220195.680155039 150427.14578372,220197.12195122 150425.12195122, 220198.661315447 150423.171302099,220200.29453926 150421.298535644,220202.017688077 150419.508163512,220203.826610682 150417.804498867, 220205.716949223 150416.191645986,220207.684149708 150414.673490372,220209.72347298 150413.253689397,220211.830006129 150411.935663483, 220213.998674333 150410.722587873,220216.22425308 150409.61738497,220218.501380756 150408.622717305,220220.824571561 150407.740981121, 220223.188228725 150406.974300596,220225.586657991 150406.324522731,220227 150406) --3d example SELECT ST_AsEWKT(ST_CurveToLine(ST_GeomFromEWKT('CIRCULARSTRING(220268 150415 1,220227 150505 2,220227 150406 3)'))); Output ------ LINESTRING(220268 150415 1,220269.95064912 150416.539364228 1.0181172856673, 220271.823415575 150418.17258804 1.03623457133459,220273.613787707 150419.895736857 1.05435185700189,....AD INFINITUM .... 220225.586657991 150406.324522731 1.32611114201132,220227 150406 3) --use only 2 segments to approximate quarter circle SELECT ST_AsText(ST_CurveToLine(ST_GeomFromText('CIRCULARSTRING(220268 150415,220227 150505,220227 150406)'),2)); st_astext ------------------------------ LINESTRING(220268 150415,220287.740300149 150448.342699654,220278.12195122 150485.87804878, 220244.779251566 150505.61834893,220207.243902439 150496,220187.50360229 150462.657300346, 220197.12195122 150425.12195122,220227 150406)
ST_DelaunayTriangles — Return a Delaunay triangulation around the given input points.
geometry ST_DelaunayTriangles(
geometry g1, float tolerance, int4 flags)
;
Return a Delaunay triangulation around the vertices of the input geometry. Output is a COLLECTION of polygons (for flags=0) or a MULTILINESTRING (for flags=1) or TIN (for flags=2). The tolerance, if any, is used to snap input vertices togheter.
Availability: 2.1.0 - requires GEOS >= 3.4.0.
This function supports 3d and will not drop the z-index.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
--la geometria originale -- ST_Union(ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) ) |
-- geometries overlaid multilinestring triangles SELECT ST_DelaunayTriangles( ST_Union(ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) )) As dtriag;
|
SELECT ST_DelaunayTriangles( ST_Union(ST_GeomFromText('POLYGON((175 150, 20 40, 50 60, 125 100, 175 150))'), ST_Buffer(ST_GeomFromText('POINT(110 170)'), 20) ),0.001,1) As dtriag;
|
-- this produces a table of 42 points that form an L shape SELECT (ST_DumpPoints(ST_GeomFromText( 'MULTIPOINT(14 14,34 14,54 14,74 14,94 14,114 14,134 14, 150 14,154 14,154 6,134 6,114 6,94 6,74 6,54 6,34 6, 14 6,10 6,8 6,7 7,6 8,6 10,6 30,6 50,6 70,6 90,6 110,6 130, 6 150,6 170,6 190,6 194,14 194,14 174,14 154,14 134,14 114, 14 94,14 74,14 54,14 34,14 14)'))).geom INTO TABLE l_shape; -- output as individual polygon triangles SELECT ST_AsText((ST_Dump(geom)).geom) As wkt FROM ( SELECT ST_DelaunayTriangles(ST_Collect(geom)) As geom FROM l_shape) As foo; ---wkt --- POLYGON((6 194,6 190,14 194,6 194)) POLYGON((14 194,6 190,14 174,14 194)) POLYGON((14 194,14 174,154 14,14 194)) POLYGON((154 14,14 174,14 154,154 14)) POLYGON((154 14,14 154,150 14,154 14)) POLYGON((154 14,150 14,154 6,154 14)) : :
|
ST_Difference — Returns a geometry that represents that part of geometry A that does not intersect with geometry B.
geometry ST_Difference(
geometry geomA, geometry geomB)
;
Returns a geometry that represents that part of geometry A that does not intersect with geometry B. One can think of this as GeometryA - ST_Intersection(A,B). If A is completely contained in B then an empty geometry collection is returned.
Note - order matters. B - A will always return a portion of B |
Eseguito dal modulo GEOS
Do not call with a GeometryCollection as an argument |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.20
This function supports 3d and will not drop the z-index. However it seems to only consider x y when doing the difference and tacks back on the Z-Index
|
|
--Safe for 2d. This is same geometries as what is shown for st_symdifference SELECT ST_AsText( ST_Difference( ST_GeomFromText('LINESTRING(50 100, 50 200)'), ST_GeomFromText('LINESTRING(50 50, 50 150)') ) ); st_astext --------- LINESTRING(50 150,50 200)
--When used in 3d doesn't quite do the right thing SELECT ST_AsEWKT(ST_Difference(ST_GeomFromEWKT('MULTIPOINT(-118.58 38.38 5,-118.60 38.329 6,-118.614 38.281 7)'), ST_GeomFromEWKT('POINT(-118.614 38.281 5)'))); st_asewkt --------- MULTIPOINT(-118.6 38.329 6,-118.58 38.38 5)
ST_Dump — Returns a set of geometry_dump (geom,path) rows, that make up a geometry g1.
geometry_dump[] ST_Dump(
geometry g1)
;
This is a set-returning function (SRF). It returns a set of geometry_dump rows, formed by a geometry (geom) and an array of integers (path). When the input geometry is a simple type (POINT,LINESTRING,POLYGON) a single record will be returned with an empty path array and the input geometry as geom. When the input geometry is a collection or multi it will return a record for each of the collection components, and the path will express the position of the component inside the collection.
ST_Dump is useful for expanding geometries. It is the reverse of a GROUP BY in that it creates new rows. For example it can be use to expand MULTIPOLYGONS into POLYGONS.
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Availability: PostGIS 1.0.0RC1. Requires PostgreSQL 7.3 or higher.
Prior to 1.3.4, this function crashes if used with geometries that contain CURVES. This is fixed in 1.3.4+ |
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This function supports 3d and will not drop the z-index.
SELECT sometable.field1, sometable.field1, (ST_Dump(sometable.the_geom)).geom AS the_geom FROM sometable; -- Break a compound curve into its constituent linestrings and circularstrings SELECT ST_AsEWKT(a.geom), ST_HasArc(a.geom) FROM ( SELECT (ST_Dump(p_geom)).geom AS geom FROM (SELECT ST_GeomFromEWKT('COMPOUNDCURVE(CIRCULARSTRING(0 0, 1 1, 1 0),(1 0, 0 1))') AS p_geom) AS b ) AS a; st_asewkt | st_hasarc -----------------------------+---------- CIRCULARSTRING(0 0,1 1,1 0) | t LINESTRING(1 0,0 1) | f (2 rows)
-- Polyhedral surface example -- Break a Polyhedral surface into its faces SELECT (a.p_geom).path[1] As path, ST_AsEWKT((a.p_geom).geom) As geom_ewkt FROM (SELECT ST_Dump(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )') ) AS p_geom ) AS a; path | geom_ewkt ------+------------------------------------------ 1 | POLYGON((0 0 0,0 0 1,0 1 1,0 1 0,0 0 0)) 2 | POLYGON((0 0 0,0 1 0,1 1 0,1 0 0,0 0 0)) 3 | POLYGON((0 0 0,1 0 0,1 0 1,0 0 1,0 0 0)) 4 | POLYGON((1 1 0,1 1 1,1 0 1,1 0 0,1 1 0)) 5 | POLYGON((0 1 0,0 1 1,1 1 1,1 1 0,0 1 0)) 6 | POLYGON((0 0 1,1 0 1,1 1 1,0 1 1,0 0 1))
-- TIN -- SELECT (g.gdump).path, ST_AsEWKT((g.gdump).geom) as wkt FROM (SELECT ST_Dump( ST_GeomFromEWKT('TIN ((( 0 0 0, 0 0 1, 0 1 0, 0 0 0 )), (( 0 0 0, 0 1 0, 1 1 0, 0 0 0 )) )') ) AS gdump ) AS g; -- result -- path | wkt ------+------------------------------------- {1} | TRIANGLE((0 0 0,0 0 1,0 1 0,0 0 0)) {2} | TRIANGLE((0 0 0,0 1 0,1 1 0,0 0 0))
ST_DumpPoints — Returns a set of geometry_dump (geom,path) rows of all points that make up a geometry.
geometry_dump[]ST_DumpPoints(
geometry geom)
;
This set-returning function (SRF) returns a set of geometry_dump
rows formed by a geometry (geom
) and an array of integers (path
).
The geom
component of geometry_dump
are all the POINT
s that make up the supplied geometry
The path
component of geometry_dump
(an integer[]
) is an index reference enumerating the POINT
s of the supplied geometry. For example, if a LINESTRING
is supplied, a path of {i}
is returned where i
is the nth
coordinate in the LINESTRING
. If a POLYGON
is supplied, a path of {i,j}
is returned where i
is the ring number (1 is outer; inner rings follow) and j
enumerates the POINT
s (again 1-based index).
Enhanced: 2.1.0 Faster speed. Reimplemented as native-C.
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Disponibilità: 1.5.0
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This function supports 3d and will not drop the z-index.
SELECT edge_id, (dp).path[1] As index, ST_AsText((dp).geom) As wktnode FROM (SELECT 1 As edge_id , ST_DumpPoints(ST_GeomFromText('LINESTRING(1 2, 3 4, 10 10)')) AS dp UNION ALL SELECT 2 As edge_id , ST_DumpPoints(ST_GeomFromText('LINESTRING(3 5, 5 6, 9 10)')) AS dp ) As foo; edge_id | index | wktnode ---------+-------+-------------- 1 | 1 | POINT(1 2) 1 | 2 | POINT(3 4) 1 | 3 | POINT(10 10) 2 | 1 | POINT(3 5) 2 | 2 | POINT(5 6) 2 | 3 | POINT(9 10)
SELECT path, ST_AsText(geom) FROM ( SELECT (ST_DumpPoints(g.geom)).* FROM (SELECT 'GEOMETRYCOLLECTION( POINT ( 0 1 ), LINESTRING ( 0 3, 3 4 ), POLYGON (( 2 0, 2 3, 0 2, 2 0 )), POLYGON (( 3 0, 3 3, 6 3, 6 0, 3 0 ), ( 5 1, 4 2, 5 2, 5 1 )), MULTIPOLYGON ( (( 0 5, 0 8, 4 8, 4 5, 0 5 ), ( 1 6, 3 6, 2 7, 1 6 )), (( 5 4, 5 8, 6 7, 5 4 )) ) )'::geometry AS geom ) AS g ) j; path | st_astext -----------+------------ {1,1} | POINT(0 1) {2,1} | POINT(0 3) {2,2} | POINT(3 4) {3,1,1} | POINT(2 0) {3,1,2} | POINT(2 3) {3,1,3} | POINT(0 2) {3,1,4} | POINT(2 0) {4,1,1} | POINT(3 0) {4,1,2} | POINT(3 3) {4,1,3} | POINT(6 3) {4,1,4} | POINT(6 0) {4,1,5} | POINT(3 0) {4,2,1} | POINT(5 1) {4,2,2} | POINT(4 2) {4,2,3} | POINT(5 2) {4,2,4} | POINT(5 1) {5,1,1,1} | POINT(0 5) {5,1,1,2} | POINT(0 8) {5,1,1,3} | POINT(4 8) {5,1,1,4} | POINT(4 5) {5,1,1,5} | POINT(0 5) {5,1,2,1} | POINT(1 6) {5,1,2,2} | POINT(3 6) {5,1,2,3} | POINT(2 7) {5,1,2,4} | POINT(1 6) {5,2,1,1} | POINT(5 4) {5,2,1,2} | POINT(5 8) {5,2,1,3} | POINT(6 7) {5,2,1,4} | POINT(5 4) (29 rows)
-- Polyhedral surface cube -- SELECT (g.gdump).path, ST_AsEWKT((g.gdump).geom) as wkt FROM (SELECT ST_DumpPoints(ST_GeomFromEWKT('POLYHEDRALSURFACE( ((0 0 0, 0 0 1, 0 1 1, 0 1 0, 0 0 0)), ((0 0 0, 0 1 0, 1 1 0, 1 0 0, 0 0 0)), ((0 0 0, 1 0 0, 1 0 1, 0 0 1, 0 0 0)), ((1 1 0, 1 1 1, 1 0 1, 1 0 0, 1 1 0)), ((0 1 0, 0 1 1, 1 1 1, 1 1 0, 0 1 0)), ((0 0 1, 1 0 1, 1 1 1, 0 1 1, 0 0 1)) )') ) AS gdump ) AS g; -- result -- path | wkt ---------+-------------- {1,1,1} | POINT(0 0 0) {1,1,2} | POINT(0 0 1) {1,1,3} | POINT(0 1 1) {1,1,4} | POINT(0 1 0) {1,1,5} | POINT(0 0 0) {2,1,1} | POINT(0 0 0) {2,1,2} | POINT(0 1 0) {2,1,3} | POINT(1 1 0) {2,1,4} | POINT(1 0 0) {2,1,5} | POINT(0 0 0) {3,1,1} | POINT(0 0 0) {3,1,2} | POINT(1 0 0) {3,1,3} | POINT(1 0 1) {3,1,4} | POINT(0 0 1) {3,1,5} | POINT(0 0 0) {4,1,1} | POINT(1 1 0) {4,1,2} | POINT(1 1 1) {4,1,3} | POINT(1 0 1) {4,1,4} | POINT(1 0 0) {4,1,5} | POINT(1 1 0) {5,1,1} | POINT(0 1 0) {5,1,2} | POINT(0 1 1) {5,1,3} | POINT(1 1 1) {5,1,4} | POINT(1 1 0) {5,1,5} | POINT(0 1 0) {6,1,1} | POINT(0 0 1) {6,1,2} | POINT(1 0 1) {6,1,3} | POINT(1 1 1) {6,1,4} | POINT(0 1 1) {6,1,5} | POINT(0 0 1) (30 rows)
-- Triangle -- SELECT (g.gdump).path, ST_AsText((g.gdump).geom) as wkt FROM (SELECT ST_DumpPoints( ST_GeomFromEWKT('TRIANGLE (( 0 0, 0 9, 9 0, 0 0 ))') ) AS gdump ) AS g; -- result -- path | wkt ------+------------ {1} | POINT(0 0) {2} | POINT(0 9) {3} | POINT(9 0) {4} | POINT(0 0)
-- TIN -- SELECT (g.gdump).path, ST_AsEWKT((g.gdump).geom) as wkt FROM (SELECT ST_DumpPoints( ST_GeomFromEWKT('TIN ((( 0 0 0, 0 0 1, 0 1 0, 0 0 0 )), (( 0 0 0, 0 1 0, 1 1 0, 0 0 0 )) )') ) AS gdump ) AS g; -- result -- path | wkt ---------+-------------- {1,1,1} | POINT(0 0 0) {1,1,2} | POINT(0 0 1) {1,1,3} | POINT(0 1 0) {1,1,4} | POINT(0 0 0) {2,1,1} | POINT(0 0 0) {2,1,2} | POINT(0 1 0) {2,1,3} | POINT(1 1 0) {2,1,4} | POINT(0 0 0) (8 rows)
ST_DumpRings — Returns a set of geometry_dump
rows, representing the exterior and interior rings of a polygon.
geometry_dump[] ST_DumpRings(
geometry a_polygon)
;
This is a set-returning function (SRF). It returns a set of geometry_dump
rows, defined as an integer[]
and a geometry
, aliased "path" and "geom" respectively. The "path" field holds the polygon ring index containing a single integer: 0 for the shell,
>0 for holes. The "geom" field contains the corresponding ring as a polygon.
Availability: PostGIS 1.1.3. Requires PostgreSQL 7.3 or higher.
This only works for POLYGON geometries. It will not work for MULTIPOLYGONS |
This function supports 3d and will not drop the z-index.
SELECT sometable.field1, sometable.field1, (ST_DumpRings(sometable.the_geom)).geom As the_geom FROM sometableOfpolys; SELECT ST_AsEWKT(geom) As the_geom, path FROM ST_DumpRings( ST_GeomFromEWKT('POLYGON((-8149064 5133092 1,-8149064 5132986 1,-8148996 5132839 1,-8148972 5132767 1,-8148958 5132508 1,-8148941 5132466 1,-8148924 5132394 1, -8148903 5132210 1,-8148930 5131967 1,-8148992 5131978 1,-8149237 5132093 1,-8149404 5132211 1,-8149647 5132310 1,-8149757 5132394 1, -8150305 5132788 1,-8149064 5133092 1), (-8149362 5132394 1,-8149446 5132501 1,-8149548 5132597 1,-8149695 5132675 1,-8149362 5132394 1))') ) as foo; path | the_geom ---------------------------------------------------------------------------------------------------------------- {0} | POLYGON((-8149064 5133092 1,-8149064 5132986 1,-8148996 5132839 1,-8148972 5132767 1,-8148958 5132508 1, | -8148941 5132466 1,-8148924 5132394 1, | -8148903 5132210 1,-8148930 5131967 1, | -8148992 5131978 1,-8149237 5132093 1, | -8149404 5132211 1,-8149647 5132310 1,-8149757 5132394 1,-8150305 5132788 1,-8149064 5133092 1)) {1} | POLYGON((-8149362 5132394 1,-8149446 5132501 1, | -8149548 5132597 1,-8149695 5132675 1,-8149362 5132394 1))
ST_FlipCoordinates — Returns a version of the given geometry with X and Y axis flipped. Useful for people who have built latitude/longitude features and need to fix them.
geometry ST_FlipCoordinates(
geometry geom)
;
Returns a version of the given geometry with X and Y axis flipped.
Disponibilità: 2.0
This method supports Circular Strings and Curves
This function supports 3d and will not drop the z-index.
This function supports M coordinates.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_RemoveRepeatedPoints — Converts a polygon or multi-polygon into a multi-point composed of randomly location points within the original areas.
geometry ST_MakeValid(
geometry input)
;
ST_Intersection — (T) Returns a geometry that represents the shared portion of geomA and geomB.
geometry ST_Intersection(
geometry geomA , geometry geomB )
;
geography ST_Intersection(
geography geogA , geography geogB )
;
Returns a geometry that represents the point set intersection of the Geometries.
In other words - that portion of geometry A and geometry B that is shared between the two geometries.
If the geometries do not share any space (are disjoint), then an empty geometry collection is returned.
ST_Intersection in conjunction with ST_Intersects is very useful for clipping geometries such as in bounding box, buffer, region queries where you only want to return that portion of a geometry that sits in a country or region of interest.
Geography: For geography this is really a thin wrapper around the geometry implementation. It first determines the best SRID that fits the bounding box of the 2 geography objects (if geography objects are within one half zone UTM but not same UTM will pick one of those) (favoring UTM or Lambert Azimuthal Equal Area (LAEA) north/south pole, and falling back on mercator in worst case scenario) and then intersection in that best fit planar spatial ref and retransforms back to WGS84 geography. |
Do not call with a |
If working with 3D geometries, you may want to use SFGCAL based ST_3DIntersection which does a proper 3D intersection for 3D geometries. Although this function works with Z-coordinate, it does an averaging of Z-Coordinate values when |
Eseguito dal modulo GEOS
This method is also provided by SFCGAL backend.
Availability: 1.5 support for geography data type was introduced.
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.18
SELECT ST_AsText(ST_Intersection('POINT(0 0)'::geometry, 'LINESTRING ( 2 0, 0 2 )'::geometry)); st_astext --------------- GEOMETRYCOLLECTION EMPTY (1 row) SELECT ST_AsText(ST_Intersection('POINT(0 0)'::geometry, 'LINESTRING ( 0 0, 0 2 )'::geometry)); st_astext --------------- POINT(0 0) (1 row) ---Clip all lines (trails) by country (here we assume country geom are POLYGON or MULTIPOLYGONS) -- NOTE: we are only keeping intersections that result in a LINESTRING or MULTILINESTRING because we don't -- care about trails that just share a point -- the dump is needed to expand a geometry collection into individual single MULT* parts -- the below is fairly generic and will work for polys, etc. by just changing the where clause SELECT clipped.gid, clipped.f_name, clipped_geom FROM (SELECT trails.gid, trails.f_name, (ST_Dump(ST_Intersection(country.the_geom, trails.the_geom))).geom As clipped_geom FROM country INNER JOIN trails ON ST_Intersects(country.the_geom, trails.the_geom)) As clipped WHERE ST_Dimension(clipped.clipped_geom) = 1 ; --For polys e.g. polygon landmarks, you can also use the sometimes faster hack that buffering anything by 0.0 -- except a polygon results in an empty geometry collection --(so a geometry collection containing polys, lines and points) -- buffered by 0.0 would only leave the polygons and dissolve the collection shell SELECT poly.gid, ST_Multi(ST_Buffer( ST_Intersection(country.the_geom, poly.the_geom), 0.0) ) As clipped_geom FROM country INNER JOIN poly ON ST_Intersects(country.the_geom, poly.the_geom) WHERE Not ST_IsEmpty(ST_Buffer(ST_Intersection(country.the_geom, poly.the_geom),0.0));
Geos is the default backend if not set. Note this is not a true intersection, compare to the same example using ST_3DIntersection.
set postgis.backend=geos; select ST_AsText(ST_Intersection(linestring, polygon)) As wkt from ST_GeomFromText('LINESTRING Z (2 2 6,1.5 1.5 7,1 1 8,0.5 0.5 8,0 0 10)') AS linestring CROSS JOIN ST_GeomFromText('POLYGON((0 0 8, 0 1 8, 1 1 8, 1 0 8, 0 0 8))') AS polygon; st_astext --------------------------------------- LINESTRING Z (1 1 8,0.5 0.5 8,0 0 10)
If your PostGIS is compiled with sfcgal support, have option of using sfcgal, but note if basically cases down both geometries to 2D before doing intersection and returns the ST_Force2D equivalent result which is a 2D geometry
set postgis.backend=sfcgal; select ST_AsText(ST_Intersection(linestring, polygon)) As wkt from ST_GeomFromText('LINESTRING Z (2 2 6,1.5 1.5 7,1 1 8,0.5 0.5 8,0 0 10)') AS linestring CROSS JOIN ST_GeomFromText('POLYGON((0 0 8, 0 1 8, 1 1 8, 1 0 8, 0 0 8))') AS polygon; wkt ---------------------------------------------- MULTILINESTRING((0.5 0.5,0 0),(1 1,0.5 0.5))
ST_3DIntersection, ST_Difference, ST_Dimension, ST_Dump, ST_Force2D, ST_SymDifference, ST_Intersects, ST_Multi
ST_LineToCurve — Converte un CIRCULARSTRING/CURVEDPOLYGON in LINESTRING/POLYGON
geometry ST_LineToCurve(
geometry geomANoncircular)
;
Converts plain LINESTRING/POLYGONS to CIRCULAR STRINGs and Curved Polygons. Note much fewer points are needed to describe the curved equivalent.
Availability: 1.2.2?
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
SELECT ST_AsText(ST_LineToCurve(foo.the_geom)) As curvedastext,ST_AsText(foo.the_geom) As non_curvedastext FROM (SELECT ST_Buffer('POINT(1 3)'::geometry, 3) As the_geom) As foo; curvedatext non_curvedastext --------------------------------------------------------------------|----------------------------------------------------------------- CURVEPOLYGON(CIRCULARSTRING(4 3,3.12132034355964 0.878679656440359, | POLYGON((4 3,3.94235584120969 2.41472903395162,3.77163859753386 1.85194970290473, 1 0,-1.12132034355965 5.12132034355963,4 3)) | 3.49440883690764 1.33328930094119,3.12132034355964 0.878679656440359, | 2.66671069905881 0.505591163092366,2.14805029709527 0.228361402466141, | 1.58527096604839 0.0576441587903094,1 0, | 0.414729033951621 0.0576441587903077,-0.148050297095264 0.228361402466137, | -0.666710699058802 0.505591163092361,-1.12132034355964 0.878679656440353, | -1.49440883690763 1.33328930094119,-1.77163859753386 1.85194970290472 | --ETC-- ,3.94235584120969 3.58527096604839,4 3)) --3D example SELECT ST_AsEWKT(ST_LineToCurve(ST_GeomFromEWKT('LINESTRING(1 2 3, 3 4 8, 5 6 4, 7 8 4, 9 10 4)'))); st_asewkt ------------------------------------ CIRCULARSTRING(1 2 3,5 6 4,9 10 4)
ST_MakeValid — Attempts to make an invalid geometry valid without losing vertices.
geometry ST_MakeValid(
geometry input)
;
The function attempts to create a valid representation of a given invalid geometry without losing any of the input vertices. Already-valid geometries are returned without further intervention.
Supported inputs are: POINTS, MULTIPOINTS, LINESTRINGS, MULTILINESTRINGS, POLYGONS, MULTIPOLYGONS and GEOMETRYCOLLECTIONS containing any mix of them.
In case of full or partial dimensional collapses, the output geometry may be a collection of lower-to-equal dimension geometries or a geometry of lower dimension.
Single polygons may become multi-geometries in case of self-intersections.
Availability: 2.0.0, requires GEOS-3.3.0
Enhanced: 2.0.1, speed improvements requires GEOS-3.3.4
Enhanced: 2.1.0 added support for GEOMETRYCOLLECTION and MULTIPOINT.
This function supports 3d and will not drop the z-index.
ST_MemUnion — Same as ST_Union, only memory-friendly (uses less memory and more processor time).
geometry ST_MemUnion(
geometry set geomfield)
;
Some useful description here.
Same as ST_Union, only memory-friendly (uses less memory and more processor time). This aggregate function works by unioning the geometries one at a time to previous result as opposed to ST_Union aggregate which first creates an array and then unions |
This function supports 3d and will not drop the z-index.
ST_MinimumBoundingCircle — Returns the smallest circle polygon that can fully contain a geometry. Default uses 48 segments per quarter circle.
geometry ST_MinimumBoundingCircle(
geometry geomA, integer num_segs_per_qt_circ=48)
;
Returns the smallest circle polygon that can fully contain a geometry.
The circle is approximated by a polygon with a default of 48 segments per quarter circle. Because the polygon is an approximation of the minimum bounding circle, some points in the input geometry may not be contained within the polygon. The approximation can be improved by increasing the number of segments, with little performance penalty. For applications where a polygonal approximation is not suitable, ST_MinimumBoundingRadius may be used. |
It is often used with MULTI and Geometry Collections. Although it is not an aggregate - you can use it in conjunction with ST_Collect to get the minimum bounding circle of a set of geometries. ST_MinimumBoundingCircle(ST_Collect(somepointfield)).
The ratio of the area of a polygon divided by the area of its Minimum Bounding Circle is often referred to as the Roeck test.
Availability: 1.4.0 - requires GEOS
SELECT d.disease_type, ST_MinimumBoundingCircle(ST_Collect(d.the_geom)) As the_geom FROM disease_obs As d GROUP BY d.disease_type;
SELECT ST_AsText(ST_MinimumBoundingCircle( ST_Collect( ST_GeomFromEWKT('LINESTRING(55 75,125 150)'), ST_Point(20, 80)), 8 )) As wktmbc; wktmbc ----------- POLYGON((135.59714732062 115,134.384753327498 102.690357210921,130.79416296937 90.8537670908995,124.963360620072 79.9451031602111,117.116420743937 70.3835792560632,107.554896839789 62.5366393799277,96.6462329091006 56.70583703063,84.8096427890789 53.115246672502,72.5000000000001 51.9028526793802,60.1903572109213 53.1152466725019,48.3537670908996 56.7058370306299,37.4451031602112 62.5366393799276,27.8835792560632 70.383579256063,20.0366393799278 79.9451031602109,14.20583703063 90.8537670908993,10.615246672502 102.690357210921,9.40285267938019 115,10.6152466725019 127.309642789079,14.2058370306299 139.1462329091,20.0366393799275 150.054896839789,27.883579256063 159.616420743937, 37.4451031602108 167.463360620072,48.3537670908992 173.29416296937,60.190357210921 176.884753327498, 72.4999999999998 178.09714732062,84.8096427890786 176.884753327498,96.6462329091003 173.29416296937,107.554896839789 167.463360620072, 117.116420743937 159.616420743937,124.963360620072 150.054896839789,130.79416296937 139.146232909101,134.384753327498 127.309642789079,135.59714732062 115))
ST_MinimumBoundingCircle — Returns the center point and radius of the smallest circle that can fully contain a geometry.
geometry_dump[]ST_DumpPoints(
geometry geom)
;
Returns a record containing the center point and radius of the smallest circle that can fully contain a geometry.
Can be used in conjunction with ST_Collect to get the minimum bounding circle of a set of geometries.
Disponibilità: 2.0
ST_Polygonize — Aggregate. Creates a GeometryCollection containing possible polygons formed from the constituent linework of a set of geometries.
geometry ST_Polygonize(
geometry set geomfield)
;
geometry ST_Polygonize(
geometry[] geom_array)
;
Creates a GeometryCollection containing possible polygons formed from the constituent linework of a set of geometries.
Geometry Collections are often difficult to deal with with third party tools, so use ST_Polygonize in conjunction with ST_Dump to dump the polygons out into individual polygons. |
Input linework must be correctly noded for this function to work properly |
Availability: 1.0.0RC1 - requires GEOS >= 2.1.0.
SELECT ST_AsEWKT(ST_Polygonize(the_geom_4269)) As geomtextrep FROM (SELECT the_geom_4269 FROM ma.suffolk_edges ORDER BY tlid LIMIT 45) As foo; geomtextrep ------------------------------------- SRID=4269;GEOMETRYCOLLECTION(POLYGON((-71.040878 42.285678,-71.040943 42.2856,-71.04096 42.285752,-71.040878 42.285678)), POLYGON((-71.17166 42.353675,-71.172026 42.354044,-71.17239 42.354358,-71.171794 42.354971,-71.170511 42.354855, -71.17112 42.354238,-71.17166 42.353675))) (1 row) --Use ST_Dump to dump out the polygonize geoms into individual polygons SELECT ST_AsEWKT((ST_Dump(foofoo.polycoll)).geom) As geomtextrep FROM (SELECT ST_Polygonize(the_geom_4269) As polycoll FROM (SELECT the_geom_4269 FROM ma.suffolk_edges ORDER BY tlid LIMIT 45) As foo) As foofoo; geomtextrep ------------------------ SRID=4269;POLYGON((-71.040878 42.285678,-71.040943 42.2856,-71.04096 42.285752, -71.040878 42.285678)) SRID=4269;POLYGON((-71.17166 42.353675,-71.172026 42.354044,-71.17239 42.354358 ,-71.171794 42.354971,-71.170511 42.354855,-71.17112 42.354238,-71.17166 42.353675)) (2 rows)
ST_Node — Node a set of linestrings.
geometry ST_Node(
geometry geom)
;
Fully node a set of linestrings using the least possible number of nodes while preserving all of the input ones.
This function supports 3d and will not drop the z-index.
Availability: 2.0.0 - requires GEOS >= 3.3.0.
Due to a bug in GEOS up to 3.3.1 this function fails to node self-intersecting lines. This is fixed with GEOS 3.3.2 or higher. |
ST_OffsetCurve — Return an offset line at a given distance and side from an input line. Useful for computing parallel lines about a center line
geometry ST_OffsetCurve(
geometry line, float signed_distance, text style_parameters='')
;
Return an offset line at a given distance and side from an input line. All points of the returned geometries are not further than the given distance from the input geometry.
For positive distance the offset will be at the left side of the input line and retain the same direction. For a negative distance it'll be at the right side and in the opposite direction.
Availability: 2.0 - requires GEOS >= 3.2, improved with GEOS >= 3.3
The optional third parameter allows specifying a list of blank-separated key=value pairs to tweak operations as follows:
'quad_segs=#' : number of segments used to approximate a quarter circle (defaults to 8).
'join=round|mitre|bevel' : join style (defaults to "round"). 'miter' is also accepted as a synonym for 'mitre'.
'mitre_limit=#.#' : mitre ratio limit (only affects mitred join style). 'miter_limit' is also accepted as a synonym for 'mitre_limit'.
Units of distance are measured in units of the spatial reference system.
The inputs can only be LINESTRINGS.
Eseguito dal modulo GEOS
This function ignores the third dimension (z) and will always give a 2-d result even when presented with a 3d-geometry. |
Compute an open buffer around roads
SELECT ST_Union( ST_OffsetCurve(f.the_geom, f.width/2, 'quad_segs=4 join=round'), ST_OffsetCurve(f.the_geom, -f.width/2, 'quad_segs=4 join=round') ) as track FROM someroadstable;
SELECT ST_AsText(ST_OffsetCurve(ST_GeomFromText( 'LINESTRING(164 16,144 16,124 16,104 16,84 16,64 16, 44 16,24 16,20 16,18 16,17 17, 16 18,16 20,16 40,16 60,16 80,16 100, 16 120,16 140,16 160,16 180,16 195)'), 15, 'quad_segs=4 join=round')); --output -- LINESTRING(164 1,18 1,12.2597485145237 2.1418070123307, 7.39339828220179 5.39339828220179, 5.39339828220179 7.39339828220179, 2.14180701233067 12.2597485145237,1 18,1 195)
|
SELECT ST_AsText(ST_OffsetCurve(geom, -15, 'quad_segs=4 join=round')) As notsocurvy FROM ST_GeomFromText( 'LINESTRING(164 16,144 16,124 16,104 16,84 16,64 16, 44 16,24 16,20 16,18 16,17 17, 16 18,16 20,16 40,16 60,16 80,16 100, 16 120,16 140,16 160,16 180,16 195)') As geom; -- notsocurvy -- LINESTRING(31 195,31 31,164 31)
|
SELECT ST_AsText(ST_OffsetCurve(ST_OffsetCurve(geom, -30, 'quad_segs=4 join=round'), -15, 'quad_segs=4 join=round')) As morecurvy FROM ST_GeomFromText( 'LINESTRING(164 16,144 16,124 16,104 16,84 16,64 16, 44 16,24 16,20 16,18 16,17 17, 16 18,16 20,16 40,16 60,16 80,16 100, 16 120,16 140,16 160,16 180,16 195)') As geom; -- morecurvy -- LINESTRING(164 31,46 31,40.2597485145236 32.1418070123307, 35.3933982822018 35.3933982822018, 32.1418070123307 40.2597485145237,31 46,31 195)
|
SELECT ST_AsText(ST_Collect( ST_OffsetCurve(geom, 15, 'quad_segs=4 join=round'), ST_OffsetCurve(ST_OffsetCurve(geom, -30, 'quad_segs=4 join=round'), -15, 'quad_segs=4 join=round') ) ) As parallel_curves FROM ST_GeomFromText( 'LINESTRING(164 16,144 16,124 16,104 16,84 16,64 16, 44 16,24 16,20 16,18 16,17 17, 16 18,16 20,16 40,16 60,16 80,16 100, 16 120,16 140,16 160,16 180,16 195)') As geom; -- parallel curves -- MULTILINESTRING((164 1,18 1,12.2597485145237 2.1418070123307, 7.39339828220179 5.39339828220179,5.39339828220179 7.39339828220179, 2.14180701233067 12.2597485145237,1 18,1 195), (164 31,46 31,40.2597485145236 32.1418070123307,35.3933982822018 35.3933982822018, 32.1418070123307 40.2597485145237,31 46,31 195))
|
SELECT ST_AsText(ST_OffsetCurve(ST_GeomFromText( 'LINESTRING(164 16,144 16,124 16,104 16,84 16,64 16, 44 16,24 16,20 16,18 16,17 17, 16 18,16 20,16 40,16 60,16 80,16 100, 16 120,16 140,16 160,16 180,16 195)'), 15, 'quad_segs=4 join=bevel')); -- output -- LINESTRING(164 1,18 1,7.39339828220179 5.39339828220179, 5.39339828220179 7.39339828220179,1 18,1 195)
|
SELECT ST_AsText(ST_Collect( ST_OffsetCurve(geom, 15, 'quad_segs=4 join=mitre mitre_limit=2.2'), ST_OffsetCurve(geom, -15, 'quad_segs=4 join=mitre mitre_limit=2.2') ) ) FROM ST_GeomFromText( 'LINESTRING(164 16,144 16,124 16,104 16,84 16,64 16, 44 16,24 16,20 16,18 16,17 17, 16 18,16 20,16 40,16 60,16 80,16 100, 16 120,16 140,16 160,16 180,16 195)') As geom; -- output -- MULTILINESTRING((164 1,11.7867965644036 1,1 11.7867965644036,1 195), (31 195,31 31,164 31))
|
ST_RemoveRepeatedPoints — Returns a version of the given geometry with duplicated points removed.
geometry ST_RemoveRepeatedPoints(
geometry geom, float8 tolerance)
;
Returns a version of the given geometry with duplicated points removed. Will actually do something only with (multi)lines, (multi)polygons and multipoints but you can safely call it with any kind of geometry. Since simplification occurs on a object-by-object basis you can also feed a GeometryCollection to this function.
If the tolerance parameter is provided, vertices within the tolerance of one another will be considered the "same" for the purposes of removal.
Availability: 2.2.0
This function supports Polyhedral surfaces.
This function supports 3d and will not drop the z-index.
ST_SharedPaths — Returns a collection containing paths shared by the two input linestrings/multilinestrings.
geometry ST_SharedPaths(
geometry lineal1, geometry lineal2)
;
Returns a collection containing paths shared by the two input geometries. Those going in the same direction are in the first element of the collection, those going in the opposite direction are in the second element. The paths themselves are given in the direction of the first geometry.
Availability: 2.0.0 requires GEOS >= 3.3.0.
SELECT ST_AsText( ST_SharedPaths( ST_GeomFromText('MULTILINESTRING((26 125,26 200,126 200,126 125,26 125), (51 150,101 150,76 175,51 150))'), ST_GeomFromText('LINESTRING(151 100,126 156.25,126 125,90 161, 76 175)') ) ) As wkt wkt ------------------------------------------------------------- GEOMETRYCOLLECTION(MULTILINESTRING((126 156.25,126 125), (101 150,90 161),(90 161,76 175)),MULTILINESTRING EMPTY)
|
-- same example but linestring orientation flipped SELECT ST_AsText( ST_SharedPaths( ST_GeomFromText('LINESTRING(76 175,90 161,126 125,126 156.25,151 100)'), ST_GeomFromText('MULTILINESTRING((26 125,26 200,126 200,126 125,26 125), (51 150,101 150,76 175,51 150))') ) ) As wkt wkt ------------------------------------------------------------- GEOMETRYCOLLECTION(MULTILINESTRING EMPTY, MULTILINESTRING((76 175,90 161),(90 161,101 150),(126 125,126 156.25)))
|
ST_ShiftLongitude — Reads every point/vertex in every component of every feature in a geometry, and if the longitude coordinate is <0, adds 360 to it. The result would be a 0-360 version of the data to be plotted in a 180 centric map
geometry ST_ShiftLongitude(
geometry geomA)
;
Reads every point/vertex in every component of every feature in a geometry, and if the longitude coordinate is <0, adds 360 to it. The result would be a 0-360 version of the data to be plotted in a 180 centric map
This is only useful for data in long lat e.g. 4326 (WGS 84 long lat) |
Pre-1.3.4 bug prevented this from working for MULTIPOINT. 1.3.4+ works with MULTIPOINT as well.
This function supports 3d and will not drop the z-index.
Enhanced: 2.0.0 support for Polyhedral surfaces and TIN was introduced.
NOTE: this function was renamed from "ST_Shift_Longitude" in 2.2.0
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
--3d points SELECT ST_AsEWKT(ST_ShiftLongitude(ST_GeomFromEWKT('SRID=4326;POINT(-118.58 38.38 10)'))) As geomA, ST_AsEWKT(ST_ShiftLongitude(ST_GeomFromEWKT('SRID=4326;POINT(241.42 38.38 10)'))) As geomb geomA geomB ---------- ----------- SRID=4326;POINT(241.42 38.38 10) SRID=4326;POINT(-118.58 38.38 10) --regular line string SELECT ST_AsText(ST_ShiftLongitude(ST_GeomFromText('LINESTRING(-118.58 38.38, -118.20 38.45)'))) st_astext ---------- LINESTRING(241.42 38.38,241.8 38.45)
ST_Simplify — Returns a "simplified" version of the given geometry using the Douglas-Peucker algorithm.
geometry ST_ConcaveHull(
geometry geomA, float target_percent, boolean allow_holes=false)
;
Returns a "simplified" version of the given geometry using the Douglas-Peucker algorithm. Will actually do something only with (multi)lines and (multi)polygons but you can safely call it with any kind of geometry. Since simplification occurs on a object-by-object basis you can also feed a GeometryCollection to this function.
The "preserve collapsed" flag will retain objects that would otherwise be too small given the tolerance. For example, a 1m long line simplified with a 10m tolerance. If the preserve flag is given, the line will not disappear. This flag is useful for rendering engines, to avoid having large numbers of very small objects disappear from a map leaving surprising gaps.
Note that returned geometry might lose its simplicity (see ST_IsSimple) |
Note topology may not be preserved and may result in invalid geometries. Use (see ST_SimplifyPreserveTopology) to preserve topology. |
Availability: 1.2.2
A circle simplified too much becomes a triangle, medium an octagon,
SELECT ST_Npoints(the_geom) As np_before, ST_NPoints(ST_Simplify(the_geom,0.1)) As np01_notbadcircle, ST_NPoints(ST_Simplify(the_geom,0.5)) As np05_notquitecircle, ST_NPoints(ST_Simplify(the_geom,1)) As np1_octagon, ST_NPoints(ST_Simplify(the_geom,10)) As np10_triangle, (ST_Simplify(the_geom,100) is null) As np100_geometrygoesaway FROM (SELECT ST_Buffer('POINT(1 3)', 10,12) As the_geom) As foo; -result np_before | np01_notbadcircle | np05_notquitecircle | np1_octagon | np10_triangle | np100_geometrygoesaway -----------+-------------------+---------------------+-------------+---------------+------------------------ 49 | 33 | 17 | 9 | 4 | t
ST_SimplifyPreserveTopology — Returns a "simplified" version of the given geometry using the Douglas-Peucker algorithm. Will avoid creating derived geometries (polygons in particular) that are invalid.
geometry ST_SimplifyPreserveTopology(
geometry geomA, float tolerance)
;
Returns a "simplified" version of the given geometry using the Douglas-Peucker algorithm. Will avoid creating derived geometries (polygons in particular) that are invalid. Will actually do something only with (multi)lines and (multi)polygons but you can safely call it with any kind of geometry. Since simplification occurs on a object-by-object basis you can also feed a GeometryCollection to this function.
Eseguito dal modulo GEOS
Requires GEOS 3.0.0+ |
Availability: 1.3.3
Same example as Simplify, but we see Preserve Topology prevents oversimplification. The circle can at most become a square.
SELECT ST_Npoints(the_geom) As np_before, ST_NPoints(ST_SimplifyPreserveTopology(the_geom,0.1)) As np01_notbadcircle, ST_NPoints(ST_SimplifyPreserveTopology(the_geom,0.5)) As np05_notquitecircle, ST_NPoints(ST_SimplifyPreserveTopology(the_geom,1)) As np1_octagon, ST_NPoints(ST_SimplifyPreserveTopology(the_geom,10)) As np10_square, ST_NPoints(ST_SimplifyPreserveTopology(the_geom,100)) As np100_stillsquare FROM (SELECT ST_Buffer('POINT(1 3)', 10,12) As the_geom) As foo; --result-- np_before | np01_notbadcircle | np05_notquitecircle | np1_octagon | np10_square | np100_stillsquare -----------+-------------------+---------------------+-------------+---------------+------------------- 49 | 33 | 17 | 9 | 5 | 5
ST_SimplifyVW — Returns a "simplified" version of the given geometry using the Visvalingam-Whyatt algorithm
geometry ST_SimplifyVW(
geometry geomA, float tolerance)
;
Returns a "simplified" version of the given geometry using the Visvalingam-Whyatt algorithm. Will actually do something only with (multi)lines and (multi)polygons but you can safely call it with any kind of geometry. Since simplification occurs on a object-by-object basis you can also feed a GeometryCollection to this function.
Note that returned geometry might lose its simplicity (see ST_IsSimple) |
Note topology may not be preserved and may result in invalid geometries. Use (see ST_SimplifyPreserveTopology) to preserve topology. |
This function handles 3D and the third dimension will affect the result. |
Availability: 2.2.0
ST_SetEffectiveArea — Sets the effective area for each vertex, storing the value in the M ordinate. A simplified geometry can then be generated by filtering on the M ordinate.
geometry ST_OffsetCurve(
geometry line, float signed_distance, text style_parameters='')
;
Sets the effective area for each vertex, using the Visvalingam-Whyatt algorithm. The effective area is stored as the M-value of the vertex. If the optional "theshold" parameter is used, a simplified geometry will be returned, containing only vertices with an effective area greater than or equal to the threshold value.
This function can be used for server-side simplification when a threshold is specified. Another option is to use a threshold value of zero. In this case, the full geometry will be returned with effective areas as M-values, which can be used by the client to simplify very quickly.
Will actually do something only with (multi)lines and (multi)polygons but you can safely call it with any kind of geometry. Since simplification occurs on a object-by-object basis you can also feed a GeometryCollection to this function.
Note that returned geometry might lose its simplicity (see ST_IsSimple) |
Note topology may not be preserved and may result in invalid geometries. Use (see ST_SimplifyPreserveTopology) to preserve topology. |
The output geometry will lose all previous information in the M-values |
This function handles 3D and the third dimension will affect the effective area |
Availability: 2.2.0
Calculating the effective area of a LineString. Because we use a threshold value of zero, all vertices in the input geometry are returned.
select ST_AsText(ST_SetEffectiveArea(geom)) all_pts, ST_AsText(ST_SetEffectiveArea(geom,30) ) thrshld_30 FROM (SELECT 'LINESTRING(5 2, 3 8, 6 20, 7 25, 10 10)'::geometry geom) As foo; -result all_pts | thrshld_30 -----------+-------------------+ LINESTRING M (5 2 3.40282346638529e+38,3 8 29,6 20 1.5,7 25 49.5,10 10 3.40282346638529e+38) | LINESTRING M (5 2 3.40282346638529e+38,7 25 49.5,10 10 3.40282346638529e+38)
ST_Split — Returns a collection of geometries resulting by splitting a geometry.
geometry ST_Split(
geometry input, geometry blade)
;
The function supports splitting a line by (multi)point, (multi)line or (multi)polygon boundary, a (multi)polygon by line. The returned geometry is always a collection.
Think of this function as the opposite of ST_Union. Theoretically applying ST_Union to the elements of the returned collection should always yield the original geometry.
Disponibilità: 2.0
Changed: 2.2.0 support for splitting a line by a multiline, a multipoint or (multi)polygon boundary was introduced.
To improve the robustness of ST_Split it may be convenient to ST_Snap the input to the blade in advance using a very low tolerance. Otherwise the internally used coordinate grid may cause tolerance problems, where coordinates of input and blade do not fall onto each other and the input is not being split correctly (see #2192). |
When a (multi)polygon is passed as as the blade, its linear component (the boundary) is used for cutting the input. |
Polygon Cut by Line
|
|
-- this creates a geometry collection consisting of the 2 halves of the polygon -- this is similar to the example we demonstrated in ST_BuildArea SELECT ST_Split(circle, line) FROM (SELECT ST_MakeLine(ST_MakePoint(10, 10),ST_MakePoint(190, 190)) As line, ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50) As circle) As foo; -- result -- GEOMETRYCOLLECTION(POLYGON((150 90,149.039264020162 80.2454838991936,146.193976625564 70.8658283817455,..), POLYGON(..))) -- To convert to individual polygons, you can use ST_Dump or ST_GeometryN SELECT ST_AsText((ST_Dump(ST_Split(circle, line))).geom) As wkt FROM (SELECT ST_MakeLine(ST_MakePoint(10, 10),ST_MakePoint(190, 190)) As line, ST_Buffer(ST_GeomFromText('POINT(100 90)'), 50) As circle) As foo; -- result -- wkt --------------- POLYGON((150 90,149.039264020162 80.2454838991936,..)) POLYGON((60.1371179574584 60.1371179574584,58.4265193848728 62.2214883490198,53.8060233744357 ..))
Multilinestring Cut by point
|
|
SELECT ST_AsText(ST_Split(mline, pt)) As wktcut FROM (SELECT ST_GeomFromText('MULTILINESTRING((10 10, 190 190), (15 15, 30 30, 100 90))') As mline, ST_Point(30,30) As pt) As foo; wktcut ------ GEOMETRYCOLLECTION( LINESTRING(10 10,30 30), LINESTRING(30 30,190 190), LINESTRING(15 15,30 30), LINESTRING(30 30,100 90) )
ST_SymDifference — Returns a geometry that represents the portions of A and B that do not intersect. It is called a symmetric difference because ST_SymDifference(A,B) = ST_SymDifference(B,A).
geometry ST_SymDifference(
geometry geomA, geometry geomB)
;
Returns a geometry that represents the portions of A and B that do not intersect. It is called a symmetric difference because ST_SymDifference(A,B) = ST_SymDifference(B,A). One can think of this as ST_Union(geomA,geomB) - ST_Intersection(A,B).
Eseguito dal modulo GEOS
Do not call with a GeometryCollection as an argument |
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.3
This method implements the SQL/MM specification. SQL-MM 3: 5.1.21
This function supports 3d and will not drop the z-index. However it seems to only consider x y when doing the difference and tacks back on the Z-Index
|
|
--Safe for 2d - symmetric difference of 2 linestrings SELECT ST_AsText( ST_SymDifference( ST_GeomFromText('LINESTRING(50 100, 50 200)'), ST_GeomFromText('LINESTRING(50 50, 50 150)') ) ); st_astext --------- MULTILINESTRING((50 150,50 200),(50 50,50 100))
--When used in 3d doesn't quite do the right thing SELECT ST_AsEWKT(ST_SymDifference(ST_GeomFromEWKT('LINESTRING(1 2 1, 1 4 2)'), ST_GeomFromEWKT('LINESTRING(1 1 3, 1 3 4)'))) st_astext ------------ MULTILINESTRING((1 3 2.75,1 4 2),(1 1 3,1 2 2.25))
ST_Subdivide — Returns a set of geometry where no geometry in the set has more than the specified number of vertices.
setof geometry ST_Subdivide(
geometry geom, integer max_vertices=256)
;
Turns a single geometry into a set in which each element has fewer than the maximum allowed number of vertices. Useful for converting excessively large polygons and other objects into small portions that fit within the database page size. Uses the same envelope clipping as ST_ClipByBox2D does, recursively subdividing the input geometry until all portions have less than the maximum vertex count. Minimum vertice count allowed is 8 and if you try to specify lower than 8, it will throw an error.
Clipping performed by the GEOS module.
Requires GEOS 3.5.0+ |
Availability: 2.2.0 requires GEOS >= 3.5.0.
-- Create a new subdivided table suitable for joining to the original CREATE TABLE subdivided_geoms AS SELECT pkey, ST_Subdivide(geom) AS geom FROM original_geoms;
SELECT row_number() OVER() As rn, ST_AsText(geom) As wkt FROM ( SELECT ST_SubDivide('POLYGON((132 10,119 23,85 35,68 29,66 28,49 42,32 56,22 64,32 110,40 119,36 150, 57 158,75 171,92 182,114 184,132 186,146 178,176 184,179 162,184 141,190 122, 190 100,185 79,186 56,186 52,178 34,168 18,147 13,132 10))'::geometry,10)) As f(geom);
rn | wkt ---+--------------------------------------------------------------------------- 1 | POLYGON((22 64,29.3913043478263 98.000000000001,106.000000000001 98.00000000001, 106.000000000001 27.5882352941173,85 35,68 29,66 28,49 42,32 56,22 64)) 2 | POLYGON((29.3913043478263 98.000000000001,32 110,40 119,36 150,57 158, 75 11,92 182,106.000000000001 183.272727272727,106.000000000001 98.000000000001, 29.913043478263 98.000000000001)) 3 | POLYGON((106.000000000001 27.5882352941173,106.000000000001 98.00000000000, 189.52380952381 98.000000000001,185 79,186 56,186 52,178 34,168 18,147 13, 132 0,119 23,106.000000000001 27.5882352941173)) 4 | POLYGON((106.000000000001 98.000000000001,106.000000000001 183.27272727272, 114 184,132 186,146 178,176 184,179 162,184 141,190 122,190 100,189.5238095238 98.000000000001,106.000000000001 98.000000000001))
|
SELECT ST_AsText(ST_SubDivide(ST_Segmentize('LINESTRING(0 0, 100 100, 150 150)'::geometry,10),8));
LINESTRING(0 0,7.07106781186547 7.07106781186547,14.1421356237309 14.1421356237309,21.2132034355964 21.2132034355964,28.2842712474619 28.2842712474619,35.3553390593274 35.3553390593274,37.499999999998 37.499999999998) LINESTRING(37.499999999998 37.499999999998,42.4264068711929 42.4264068711929,49.4974746830583 49.4974746830583,56.5685424949238 56.5685424949238,63.6396103067893 63.6396103067893,70.7106781186548 70.7106781186548,74.999999999998 74.999999999998) LINESTRING(74.999999999998 74.999999999998,77.7817459305202 77.7817459305202,84.8528137423857 84.8528137423857,91.9238815542512 91.9238815542512,98.9949493661167 98.9949493661167,100 100,107.071067811865 107.071067811865,112.499999999998 112.499999999998) LINESTRING(112.499999999998 112.499999999998,114.142135623731 114.142135623731,121.213203435596 121.213203435596,128.284271247462 128.284271247462,135.355339059327 135.355339059327,142.426406871193 142.426406871193,149.497474683058 149.497474683058,149.999999999998 149.999999999998)
|
ST_SwapOrdinates — Returns a version of the given geometry with given ordinate values swapped.
geometry ST_SwapOrdinates(
geometry geom, cstring ords)
;
Returns a version of the given geometry with given ordinates swapped.
The ords
parameter is a 2-characters string naming the ordinates to swap. Valid names are: x,y,z and m.
Availability: 2.2.0
This method supports Circular Strings and Curves
This function supports 3d and will not drop the z-index.
This function supports M coordinates.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
ST_Union — Returns a geometry that represents the point set union of the Geometries.
geometry ST_Union(
geometry set g1field)
;
geometry ST_Union(
geometry g1, geometry g2)
;
geometry ST_Union(
geometry[] g1_array)
;
Output type can be a MULTI*, single geometry, or Geometry Collection. Comes in 2 variants. Variant 1 unions 2 geometries resulting in a new geometry with no intersecting regions. Variant 2 is an aggregate function that takes a set of geometries and unions them into a single ST_Geometry resulting in no intersecting regions.
Aggregate version: This function returns a MULTI geometry or NON-MULTI geometry from a set of geometries. The ST_Union() function is an "aggregate" function in the terminology of PostgreSQL. That means that it operates on rows of data, in the same way the SUM() and AVG() functions do and like most aggregates, it also ignores NULL geometries.
Non-Aggregate version: This function returns a geometry being a union of two input geometries. Output type can be a MULTI*, NON-MULTI or GEOMETRYCOLLECTION. If any are NULL, then NULL is returned.
ST_Collect and ST_Union are often interchangeable. ST_Union is in general orders of magnitude slower than ST_Collect because it tries to dissolve boundaries and reorder geometries to ensure that a constructed Multi* doesn't have intersecting regions. |
Eseguito dal modulo GEOS
NOTE: this function was formerly called GeomUnion(), which was renamed from "Union" because UNION is an SQL reserved word.
Availability: 1.4.0 - ST_Union was enhanced. ST_Union(geomarray) was introduced and also faster aggregate collection in PostgreSQL. If you are using GEOS 3.1.0+ ST_Union will use the faster Cascaded Union algorithm described in http://blog.cleverelephant.ca/2009/01/must-faster-unions-in-postgis-14.html
This method implements the OpenGIS Simple Features Implementation Specification for SQL 1.1. s2.1.1.3
Aggregate version is not explicitly defined in OGC SPEC. |
This method implements the SQL/MM specification. SQL-MM 3: 5.1.19 the z-index (elevation) when polygons are involved.
Aggregate example
SELECT stusps, ST_Multi(ST_Union(f.the_geom)) as singlegeom FROM sometable As f GROUP BY stusps
Non-Aggregate example
SELECT ST_AsText(ST_Union(ST_GeomFromText('POINT(1 2)'), ST_GeomFromText('POINT(-2 3)') ) ) st_astext ---------- MULTIPOINT(-2 3,1 2) SELECT ST_AsText(ST_Union(ST_GeomFromText('POINT(1 2)'), ST_GeomFromText('POINT(1 2)') ) ); st_astext ---------- POINT(1 2) --3d example - sort of supports 3d (and with mixed dimensions!) SELECT ST_AsEWKT(st_union(the_geom)) FROM (SELECT ST_GeomFromEWKT('POLYGON((-7 4.2,-7.1 4.2,-7.1 4.3, -7 4.2))') as the_geom UNION ALL SELECT ST_GeomFromEWKT('POINT(5 5 5)') as the_geom UNION ALL SELECT ST_GeomFromEWKT('POINT(-2 3 1)') as the_geom UNION ALL SELECT ST_GeomFromEWKT('LINESTRING(5 5 5, 10 10 10)') as the_geom ) as foo; st_asewkt --------- GEOMETRYCOLLECTION(POINT(-2 3 1),LINESTRING(5 5 5,10 10 10),POLYGON((-7 4.2 5,-7.1 4.2 5,-7.1 4.3 5,-7 4.2 5))); --3d example not mixing dimensions SELECT ST_AsEWKT(st_union(the_geom)) FROM (SELECT ST_GeomFromEWKT('POLYGON((-7 4.2 2,-7.1 4.2 3,-7.1 4.3 2, -7 4.2 2))') as the_geom UNION ALL SELECT ST_GeomFromEWKT('POINT(5 5 5)') as the_geom UNION ALL SELECT ST_GeomFromEWKT('POINT(-2 3 1)') as the_geom UNION ALL SELECT ST_GeomFromEWKT('LINESTRING(5 5 5, 10 10 10)') as the_geom ) as foo; st_asewkt --------- GEOMETRYCOLLECTION(POINT(-2 3 1),LINESTRING(5 5 5,10 10 10),POLYGON((-7 4.2 2,-7.1 4.2 3,-7.1 4.3 2,-7 4.2 2))) --Examples using new Array construct SELECT ST_Union(ARRAY(SELECT the_geom FROM sometable)); SELECT ST_AsText(ST_Union(ARRAY[ST_GeomFromText('LINESTRING(1 2, 3 4)'), ST_GeomFromText('LINESTRING(3 4, 4 5)')])) As wktunion; --wktunion--- MULTILINESTRING((3 4,4 5),(1 2,3 4))
ST_UnaryUnion — Like ST_Union, but working at the geometry component level.
geometry ST_UnaryUnion(
geometry geom)
;
Unlike ST_Union, ST_UnaryUnion does dissolve boundaries between components of a multipolygon (invalid) and does perform union between the components of a geometrycollection. Each components of the input geometry is assumed to be valid, so you won't get a valid multipolygon out of a bow-tie polygon (invalid).
You may use this function to node a set of linestrings. You may mix ST_UnaryUnion with ST_Collect to fine-tune how many geometries at once you want to dissolve to be nice on both memory size and CPU time, finding the balance between ST_Union and ST_MemUnion.
This function supports 3d and will not drop the z-index.
Availability: 2.0.0 - requires GEOS >= 3.3.0.
ST_Voronoi — Computes a Voronoi diagram from the vertices of a geometry.
geometry ST_OffsetCurve(
geometry line, float signed_distance, text style_parameters='')
;
ST_Voronoi computes a two-dimensional Voronoi diagram from the vertices of the supplied geometry. By default, the result will be a GeometryCollection of Polygons that covers an envelope larger than the extent of the input vertices.
Optional parameters:
'clip' : If a geometry is supplied as the "clip" parameter, the diagram will be extended to cover the envelope of the "clip" geometry, unless that envelope is smaller than the default envelope. (default = NULL)
'tolerance' : The distance within which vertices will be considered equivalent. Robustness of the algorithm can be improved by supplying a nonzero tolerance distance. (default = 0.0)
'return_polygons' : if true, the result of ST_Voronoi will be a GeometryCollection of Polygons. If false, the result will be a MultiLineString. (default = true)
Availability: 2.3.0 - requires GEOS >= 3.5.0.
SELECT ST_Voronoi(geom) As geom FROM (SELECT 'MULTIPOINT (50 30, 60 30, 100 100,10 150, 110 120)'::geometry As geom ) As g;
-- ST_AsText output GEOMETRYCOLLECTION(POLYGON((-110 43.3333333333333,-110 270,100.5 270,59.3478260869565 132.826086956522,36.8181818181818 92.2727272727273,-110 43.3333333333333)), POLYGON((55 -90,-110 -90,-110 43.3333333333333,36.8181818181818 92.2727272727273,55 79.2857142857143,55 -90)), POLYGON((230 47.5,230 -20.7142857142857,55 79.2857142857143,36.8181818181818 92.2727272727273,59.3478260869565 132.826086956522,230 47.5)),POLYGON((230 -20.7142857142857,230 -90,55 -90,55 79.2857142857143,230 -20.7142857142857)), POLYGON((100.5 270,230 270,230 47.5,59.3478260869565 132.826086956522,100.5 270)))
|
SELECT ST_Voronoi(geom, null,30) As geom FROM (SELECT 'MULTIPOINT (50 30, 60 30, 100 100,10 150, 110 120)'::geometry As geom ) As g;
-- ST_AsText output GEOMETRYCOLLECTION(POLYGON((-110 43.3333333333333,-110 270,100.5 270,59.3478260869565 132.826086956522,36.8181818181818 92.2727272727273,-110 43.3333333333333)), POLYGON((230 47.5,230 -45.7142857142858,36.8181818181818 92.2727272727273,59.3478260869565 132.826086956522,230 47.5)),POLYGON((230 -45.7142857142858,230 -90,-110 -90,-110 43.3333333333333,36.8181818181818 92.2727272727273,230 -45.7142857142858)), POLYGON((100.5 270,230 270,230 47.5,59.3478260869565 132.826086956522,100.5 270)))
|
SELECT ST_Voronoi(geom, null,30,false) As geom FROM (SELECT 'MULTIPOINT (50 30, 60 30, 100 100,10 150, 110 120)'::geometry As geom ) As g
-- ST_AsText output MULTILINESTRING((135.555555555556 270,36.8181818181818 92.2727272727273),(36.8181818181818 92.2727272727273,-110 43.3333333333333),(230 -45.7142857142858,36.8181818181818 92.2727272727273))
|
ST_LineInterpolatePoint — Returns a point interpolated along a line. Second argument is a float8 between 0 and 1 representing fraction of total length of linestring the point has to be located.
geometry ST_LineInterpolatePoint(
geometry a_linestring, float8 a_fraction)
;
Returns a point interpolated along a line. First argument must be a LINESTRING. Second argument is a float8 between 0 and 1 representing fraction of total linestring length the point has to be located.
See ST_LineLocatePoint for computing the line location nearest to a Point.
Since release 1.1.1 this function also interpolates M and Z values (when present), while prior releases set them to 0.0. |
Availability: 0.8.2, Z and M supported added in 1.1.1
Changed: 2.1.0. Up to 2.0.x this was called ST_Line_Interpolate_Point.
This function supports 3d and will not drop the z-index.
--Return point 20% along 2d line SELECT ST_AsEWKT(ST_LineInterpolatePoint(the_line, 0.20)) FROM (SELECT ST_GeomFromEWKT('LINESTRING(25 50, 100 125, 150 190)') as the_line) As foo; st_asewkt ---------------- POINT(51.5974135047432 76.5974135047432)
--Return point mid-way of 3d line SELECT ST_AsEWKT(ST_LineInterpolatePoint(the_line, 0.5)) FROM (SELECT ST_GeomFromEWKT('LINESTRING(1 2 3, 4 5 6, 6 7 8)') as the_line) As foo; st_asewkt -------------------- POINT(3.5 4.5 5.5) --find closest point on a line to a point or other geometry SELECT ST_AsText(ST_LineInterpolatePoint(foo.the_line, ST_LineLocatePoint(foo.the_line, ST_GeomFromText('POINT(4 3)')))) FROM (SELECT ST_GeomFromText('LINESTRING(1 2, 4 5, 6 7)') As the_line) As foo; st_astext ---------------- POINT(3 4)
ST_LineLocatePoint — Returns a float between 0 and 1 representing the location of the closest point on LineString to the given Point, as a fraction of total 2d line length.
float8 ST_LineLocatePoint(
geometry a_linestring, geometry a_point)
;
Returns a float between 0 and 1 representing the location of the closest point on LineString to the given Point, as a fraction of total 2d line length.
You can use the returned location to extract a Point (ST_LineInterpolatePoint) or a substring (ST_LineSubstring).
This is useful for approximating numbers of addresses
Disponibilità: 1.1.0
Changed: 2.1.0. Up to 2.0.x this was called ST_Line_Locate_Point.
--Rough approximation of finding the street number of a point along the street --Note the whole foo thing is just to generate dummy data that looks --like house centroids and street --We use ST_DWithin to exclude --houses too far away from the street to be considered on the street SELECT ST_AsText(house_loc) As as_text_house_loc, startstreet_num + CAST( (endstreet_num - startstreet_num) * ST_LineLocatePoint(street_line, house_loc) As integer) As street_num FROM (SELECT ST_GeomFromText('LINESTRING(1 2, 3 4)') As street_line, ST_MakePoint(x*1.01,y*1.03) As house_loc, 10 As startstreet_num, 20 As endstreet_num FROM generate_series(1,3) x CROSS JOIN generate_series(2,4) As y) As foo WHERE ST_DWithin(street_line, house_loc, 0.2); as_text_house_loc | street_num -------------------+------------ POINT(1.01 2.06) | 10 POINT(2.02 3.09) | 15 POINT(3.03 4.12) | 20 --find closest point on a line to a point or other geometry SELECT ST_AsText(ST_LineInterpolatePoint(foo.the_line, ST_LineLocatePoint(foo.the_line, ST_GeomFromText('POINT(4 3)')))) FROM (SELECT ST_GeomFromText('LINESTRING(1 2, 4 5, 6 7)') As the_line) As foo; st_astext ---------------- POINT(3 4)
ST_LineSubstring — Return a linestring being a substring of the input one starting and ending at the given fractions of total 2d length. Second and third arguments are float8 values between 0 and 1.
geometry ST_LineSubstring(
geometry a_linestring, float8 startfraction, float8 endfraction)
;
Return a linestring being a substring of the input one starting and ending at the given fractions of total 2d length. Second and third arguments are float8 values between 0 and 1. This only works with LINESTRINGs. To use with contiguous MULTILINESTRINGs use in conjunction with ST_LineMerge.
If 'start' and 'end' have the same value this is equivalent to ST_LineInterpolatePoint.
See ST_LineLocatePoint for computing the line location nearest to a Point.
Since release 1.1.1 this function also interpolates M and Z values (when present), while prior releases set them to unspecified values. |
Availability: 1.1.0, Z and M supported added in 1.1.1
Changed: 2.1.0. Up to 2.0.x this was called ST_Line_Substring.
This function supports 3d and will not drop the z-index.
--Return the approximate 1/3 mid-range part of a linestring SELECT ST_AsText(ST_Line_SubString(ST_GeomFromText('LINESTRING(25 50, 100 125, 150 190)'), 0.333, 0.666)); st_astext ------------------------------------------------------------------------------------------------ LINESTRING(69.2846934853974 94.2846934853974,100 125,111.700356260683 140.210463138888) --The below example simulates a while loop in --SQL using PostgreSQL generate_series() to cut all --linestrings in a table to 100 unit segments -- of which no segment is longer than 100 units -- units are measured in the SRID units of measurement -- It also assumes all geometries are LINESTRING or contiguous MULTILINESTRING --and no geometry is longer than 100 units*10000 --for better performance you can reduce the 10000 --to match max number of segments you expect SELECT field1, field2, ST_LineSubstring(the_geom, 100.00*n/length, CASE WHEN 100.00*(n+1) < length THEN 100.00*(n+1)/length ELSE 1 END) As the_geom FROM (SELECT sometable.field1, sometable.field2, ST_LineMerge(sometable.the_geom) AS the_geom, ST_Length(sometable.the_geom) As length FROM sometable ) AS t CROSS JOIN generate_series(0,10000) AS n WHERE n*100.00/length < 1;
ST_LocateAlong — Return a derived geometry collection value with elements that match the specified measure. Polygonal elements are not supported.
geometry ST_LocateAlong(
geometry ageom_with_measure, float8 a_measure, float8 offset)
;
Return a derived geometry collection value with elements that match the specified measure. Polygonal elements are not supported.
If an offset is provided, the resultant will be offset to the left or right of the input line by the specified number of units. A positive offset will be to the left, and a negative one to the right.
Semantic is specified by: ISO/IEC CD 13249-3:200x(E) - Text for Continuation CD Editing Meeting
Availability: 1.1.0 by old name ST_Locate_Along_Measure.
Changed: 2.0.0 in prior versions this used to be called ST_Locate_Along_Measure. The old name has been deprecated and will be removed in the future but is still available.
Use this function only for geometries with an M component |
This function supports M coordinates.
SELECT ST_AsText(the_geom) FROM (SELECT ST_LocateAlong( ST_GeomFromText('MULTILINESTRINGM((1 2 3, 3 4 2, 9 4 3), (1 2 3, 5 4 5))'),3) As the_geom) As foo; st_asewkt ----------------------------------------------------------- MULTIPOINT M (1 2 3) --Le collezioni di geometrie sono bestie difficili, per cui ne facciamo un dump --per renderle più digeribili SELECT ST_AsText((ST_Dump(the_geom)).geom) FROM (SELECT ST_LocateAlong( ST_GeomFromText('MULTILINESTRINGM((1 2 3, 3 4 2, 9 4 3), (1 2 3, 5 4 5))'),3) As the_geom) As foo; st_asewkt --------------- POINTM(1 2 3) POINTM(9 4 3) POINTM(1 2 3)
ST_LocateBetween — Return a derived geometry collection value with elements that match the specified range of measures inclusively. Polygonal elements are not supported.
geometry ST_LocateBetween(
geometry geomA, float8 measure_start, float8 measure_end, float8 offset)
;
Return a derived geometry collection value with elements that match the specified range of measures inclusively. Polygonal elements are not supported.
Semantic is specified by: ISO/IEC CD 13249-3:200x(E) - Text for Continuation CD Editing Meeting
Availability: 1.1.0 by old name ST_Locate_Between_Measures.
Changed: 2.0.0 - in prior versions this used to be called ST_Locate_Between_Measures. The old name has been deprecated and will be removed in the future but is still available for backward compatibility.
This function supports M coordinates.
SELECT ST_AsText(the_geom) FROM (SELECT ST_LocateBetween( ST_GeomFromText('MULTILINESTRING M ((1 2 3, 3 4 2, 9 4 3), (1 2 3, 5 4 5))'),1.5, 3) As the_geom) As foo; st_asewkt ------------------------------------------------------------------------ GEOMETRYCOLLECTION M (LINESTRING M (1 2 3,3 4 2,9 4 3),POINT M (1 2 3)) --Geometry collections are difficult animals so dump them --to make them more digestable SELECT ST_AsText((ST_Dump(the_geom)).geom) FROM (SELECT ST_LocateBetween( ST_GeomFromText('MULTILINESTRING M ((1 2 3, 3 4 2, 9 4 3), (1 2 3, 5 4 5))'),1.5, 3) As the_geom) As foo; st_asewkt -------------------------------- LINESTRING M (1 2 3,3 4 2,9 4 3) POINT M (1 2 3)
ST_LocateBetweenElevations — Return a derived geometry (collection) value with elements that intersect the specified range of elevations inclusively. Only 3D, 4D LINESTRINGS and MULTILINESTRINGS are supported.
geometry ST_LocateBetweenElevations(
geometry geom_mline, float8 elevation_start, float8 elevation_end)
;
Return a derived geometry (collection) value with elements that intersect the specified range of elevations inclusively. Only 3D, 3DM LINESTRINGS and MULTILINESTRINGS are supported.
Disponibilità: 1.4.0
This function supports 3d and will not drop the z-index.
SELECT ST_AsEWKT(ST_LocateBetweenElevations( ST_GeomFromEWKT('LINESTRING(1 2 3, 4 5 6)'),2,4)) As ewelev; ewelev ---------------------------------------------------------------- MULTILINESTRING((1 2 3,2 3 4)) SELECT ST_AsEWKT(ST_LocateBetweenElevations( ST_GeomFromEWKT('LINESTRING(1 2 6, 4 5 -1, 7 8 9)'),6,9)) As ewelev; ewelev ---------------------------------------------------------------- GEOMETRYCOLLECTION(POINT(1 2 6),LINESTRING(6.1 7.1 6,7 8 9)) --Geometry collections are difficult animals so dump them --to make them more digestable SELECT ST_AsEWKT((ST_Dump(the_geom)).geom) FROM (SELECT ST_LocateBetweenElevations( ST_GeomFromEWKT('LINESTRING(1 2 6, 4 5 -1, 7 8 9)'),6,9) As the_geom) As foo; st_asewkt -------------------------------- POINT(1 2 6) LINESTRING(6.1 7.1 6,7 8 9)
ST_InterpolatePoint — Return the value of the measure dimension of a geometry at the point closed to the provided point.
float8 ST_InterpolatePoint(
geometry line, geometry point)
;
Return the value of the measure dimension of a geometry at the point closed to the provided point.
Availability: 2.0.0
This function supports 3d and will not drop the z-index.
ST_AddMeasure — Return a derived geometry with measure elements linearly interpolated between the start and end points.
geometry ST_AddMeasure(
geometry geom_mline, float8 measure_start, float8 measure_end)
;
Return a derived geometry with measure elements linearly interpolated between the start and end points. If the geometry has no measure dimension, one is added. If the geometry has a measure dimension, it is over-written with new values. Only LINESTRINGS and MULTILINESTRINGS are supported.
Disponibilità: 1.5.0
This function supports 3d and will not drop the z-index.
SELECT ST_AsText(ST_AddMeasure( ST_GeomFromEWKT('LINESTRING(1 0, 2 0, 4 0)'),1,4)) As ewelev; ewelev -------------------------------- LINESTRINGM(1 0 1,2 0 2,4 0 4) SELECT ST_AsText(ST_AddMeasure( ST_GeomFromEWKT('LINESTRING(1 0 4, 2 0 4, 4 0 4)'),10,40)) As ewelev; ewelev ---------------------------------------- LINESTRING(1 0 4 10,2 0 4 20,4 0 4 40) SELECT ST_AsText(ST_AddMeasure( ST_GeomFromEWKT('LINESTRINGM(1 0 4, 2 0 4, 4 0 4)'),10,40)) As ewelev; ewelev ---------------------------------------- LINESTRINGM(1 0 10,2 0 20,4 0 40) SELECT ST_AsText(ST_AddMeasure( ST_GeomFromEWKT('MULTILINESTRINGM((1 0 4, 2 0 4, 4 0 4),(1 0 4, 2 0 4, 4 0 4))'),10,70)) As ewelev; ewelev ----------------------------------------------------------------- MULTILINESTRINGM((1 0 10,2 0 20,4 0 40),(1 0 40,2 0 50,4 0 70))
true
if the geometry is a valid trajectory. ST_IsValidTrajectory — Returns true
if the geometry is a valid trajectory.
boolean ST_IsValidTrajectory(
geometry line)
;
Tell if a geometry encodes a valid trajectory. Valid trajectories are encoded as LINESTRING with M value growing from each vertex to the next.
Valid trajectories are expected as input to some spatio-temporal queries like ST_ClosestPointOfApproach
Availability: 2.2.0
This function supports 3d and will not drop the z-index.
-- A valid trajectory SELECT ST_IsValidTrajectory(ST_MakeLine( ST_MakePointM(0,0,1), ST_MakePointM(0,1,2)) ); t -- An invalid trajectory SELECT ST_IsValidTrajectory(ST_MakeLine(ST_MakePointM(0,0,1), ST_MakePointM(0,1,0))); NOTICE: Measure of vertex 1 (0) not bigger than measure of vertex 0 (1) st_isvalidtrajectory ---------------------- f
ST_ClosestPointOfApproach — Returns the measure at which points interpolated along two lines are closest.
float8 ST_ClosestPointOfApproach(
geometry track1, geometry track2)
;
Returns the smallest measure at which point interpolated along the given lines are at the smallest distance. Inputs must be valid trajectories as checked by ST_IsValidTrajectory. Null is returned if the trajectories do not overlap on the M range.
See ST_LocateAlong for getting the actual points at the given measure.
Availability: 2.2.0
This function supports 3d and will not drop the z-index.
-- Return the time in which two objects moving between 10:00 and 11:00 -- are closest to each other and their distance at that point WITH inp AS ( SELECT ST_AddMeasure('LINESTRING Z (0 0 0, 10 0 5)'::geometry, extract(epoch from '2015-05-26 10:00'::timestamptz), extract(epoch from '2015-05-26 11:00'::timestamptz) ) a, ST_AddMeasure('LINESTRING Z (0 2 10, 12 1 2)'::geometry, extract(epoch from '2015-05-26 10:00'::timestamptz), extract(epoch from '2015-05-26 11:00'::timestamptz) ) b ), cpa AS ( SELECT ST_ClosestPointOfApproach(a,b) m FROM inp ), points AS ( SELECT ST_Force3DZ(ST_GeometryN(ST_LocateAlong(a,m),1)) pa, ST_Force3DZ(ST_GeometryN(ST_LocateAlong(b,m),1)) pb FROM inp, cpa ) SELECT to_timestamp(m) t, ST_Distance(pa,pb) distance FROM points, cpa; t | distance -------------------------------+------------------ 2015-05-26 10:45:31.034483+02 | 1.96036833151395
ST_DistanceCPA — Returns the distance between closest points of approach in two trajectories.
float8 ST_DistanceCPA(
geometry track1, geometry track2)
;
Returns the minimum distance two moving objects have ever been each-other. Inputs must be valid trajectories as checked by ST_IsValidTrajectory. Null is returned if the trajectories do not overlap on the M range.
Availability: 2.2.0
This function supports 3d and will not drop the z-index.
-- Return the minimum distance of two objects moving between 10:00 and 11:00 WITH inp AS ( SELECT ST_AddMeasure('LINESTRING Z (0 0 0, 10 0 5)'::geometry, extract(epoch from '2015-05-26 10:00'::timestamptz), extract(epoch from '2015-05-26 11:00'::timestamptz) ) a, ST_AddMeasure('LINESTRING Z (0 2 10, 12 1 2)'::geometry, extract(epoch from '2015-05-26 10:00'::timestamptz), extract(epoch from '2015-05-26 11:00'::timestamptz) ) b ) SELECT ST_DistanceCPA(a,b) distance FROM inp; distance ------------------ 1.96036833151395
ST_CPAWithin — Returns true if the trajectories' closest points of approach are within the specified distance.
float8 ST_CPAWithin(
geometry track1, geometry track2, float8 maxdist)
;
Checks whether two moving objects have ever been within the specified max distance.
Inputs must be valid trajectories as checked by ST_IsValidTrajectory. False is returned if the trajectories do not overlap on the M range.
Availability: 2.2.0
This function supports 3d and will not drop the z-index.
WITH inp AS ( SELECT ST_AddMeasure('LINESTRING Z (0 0 0, 10 0 5)'::geometry, extract(epoch from '2015-05-26 10:00'::timestamptz), extract(epoch from '2015-05-26 11:00'::timestamptz) ) a, ST_AddMeasure('LINESTRING Z (0 2 10, 12 1 2)'::geometry, extract(epoch from '2015-05-26 10:00'::timestamptz), extract(epoch from '2015-05-26 11:00'::timestamptz) ) b ) SELECT ST_CPAWithin(a,b,2), ST_DistanceCPA(a,b) distance FROM inp; st_cpawithin | distance --------------+------------------ t | 1.96521473776207
This module and associated pl/pgsql functions have been implemented to provide long locking support required by Web Feature Service specification.
Users must use serializable transaction level otherwise locking mechanism would break. |
AddAuth — Add an authorization token to be used in current transaction.
boolean AddAuth(
text auth_token)
;
Add an authorization token to be used in current transaction.
Creates/adds to a temp table called temp_lock_have_table the current transaction identifier and authorization token key.
Availability: 1.1.3
CheckAuth — Creates trigger on a table to prevent/allow updates and deletes of rows based on authorization token.
integer CheckAuth(
text a_schema_name, text a_table_name, text a_key_column_name)
;
integer CheckAuth(
text a_table_name, text a_key_column_name)
;
Creates trigger on a table to prevent/allow updates and deletes of rows based on authorization token. Identify rows using <rowid_col> column.
If a_schema_name is not passed in, then searches for table in current schema.
If an authorization trigger already exists on this table function errors. If Transaction support is not enabled, function throws an exception. |
Availability: 1.1.3
DisableLongTransactions — Disable long transaction support. This function removes the long transaction support metadata tables, and drops all triggers attached to lock-checked tables.
text DisableLongTransactions(
)
;
Disable long transaction support. This function removes the long transaction support metadata tables, and drops all triggers attached to lock-checked tables.
Drops meta table called authorization_table
and a view called authorized_tables
and all triggers called checkauthtrigger
Availability: 1.1.3
EnableLongTransactions — Enable long transaction support. This function creates the required metadata tables, needs to be called once before using the other functions in this section. Calling it twice is harmless.
text EnableLongTransactions(
)
;
Enable long transaction support. This function creates the required metadata tables, needs to be called once before using the other functions in this section. Calling it twice is harmless.
Creates a meta table called authorization_table
and a view called authorized_tables
Availability: 1.1.3
LockRow — Set lock/authorization for specific row in table
integer LockRow(
text a_schema_name, text a_table_name, text a_row_key, text an_auth_token, timestamp expire_dt)
;
integer LockRow(
text a_table_name, text a_row_key, text an_auth_token, timestamp expire_dt)
;
integer LockRow(
text a_table_name, text a_row_key, text an_auth_token)
;
Set lock/authorization for specific row in table <authid> is a text value, <expires> is a timestamp defaulting to now()+1hour. Returns 1 if lock has been assigned, 0 otherwise (already locked by other auth)
Availability: 1.1.3
UnlockRows — Remove all locks held by specified authorization id. Returns the number of locks released.
integer UnlockRows(
text auth_token)
;
Remove all locks held by specified authorization id. Returns the number of locks released.
Availability: 1.1.3
ST_Accum — Aggregate. Constructs an array of geometries.
geometry[] ST_Accum(
geometry set geomfield)
;
Aggregate. Constructs an array of geometries.
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT (ST_Accum(the_geom)) As all_em, ST_AsText((ST_Accum(the_geom))[1]) As grabone, (ST_Accum(the_geom))[2:4] as grab_rest FROM (SELECT ST_MakePoint(a*CAST(random()*10 As integer), a*CAST(random()*10 As integer), a*CAST(random()*10 As integer)) As the_geom FROM generate_series(1,4) a) As foo; all_em|grabone | grab_rest -------------------------------------------------------------------------------+ {0101000080000000000000144000000000000024400000000000001040: 0101000080000000000 00018400000000000002C400000000000003040: 0101000080000000000000354000000000000038400000000000001840: 010100008000000000000040400000000000003C400000000000003040} | POINT(5 10) | {010100008000000000000018400000000000002C400000000000003040: 0101000080000000000000354000000000000038400000000000001840: 010100008000000000000040400000000000003C400000000000003040} (1 row)
Box2D — Returns a BOX2D representing the maximum extents of the geometry.
box2d Box2D(
geometry geomA)
;
Returns a BOX2D representing the maximum extents of the geometry.
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
Box3D — Returns a BOX3D representing the maximum extents of the geometry.
box3d Box3D(
geometry geomA)
;
Returns a BOX3D representing the maximum extents of the geometry.
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
This function supports 3d and will not drop the z-index.
ST_EstimatedExtent — Return the 'estimated' extent of the given spatial table. The estimated is taken from the geometry column's statistics. The current schema will be used if not specified.
geometry ST_Expand(
geometry g1, float units_to_expand)
;
box2d ST_Expand(
box2d g1, float units_to_expand)
;
box3d ST_Expand(
box3d g1, float units_to_expand)
;
Return the 'estimated' extent of the given spatial table. The estimated is taken from the geometry column's statistics. The current schema will be used if not specified. The default behavior is to also use statistics collected from children tables (tables with INHERITS) if available. If 'parent_ony' is set to TRUE, only statistics for the given table are used and children tables are ignored.
For PostgreSQL>=8.0.0 statistics are gathered by VACUUM ANALYZE and resulting extent will be about 95% of the real one.
In absence of statistics (empty table or no ANALYZE called) this function returns NULL. Prior to version 1.5.4 an exception was thrown instead. |
For PostgreSQL<8.0.0 statistics are gathered by update_geometry_stats() and resulting extent will be exact.
Availability: 1.0.0
Modificato nella versione 2.1.0. Sino alla versione 2.0.x questa funzione si chiamava ST_Estimated_Extent.
This method supports Circular Strings and Curves
ST_Expand — Returns bounding box expanded in all directions from the bounding box of the input geometry. Uses double-precision
geometry ST_Expand(
geometry g1, float units_to_expand)
;
box2d ST_Expand(
box2d g1, float units_to_expand)
;
box3d ST_Expand(
box3d g1, float units_to_expand)
;
This function returns a bounding box expanded in all directions from the bounding box of the input geometry, by an amount specified in the second argument. Uses double-precision. Very useful for distance() queries, or bounding box queries to add an index filter to the query.
There are 3 variants of this. The one that takes a geometry will return a POLYGON geometry representation of the bounding box and is the most commonly used variant.
ST_Expand is similar in concept to ST_Buffer except while buffer expands the geometry in all directions, ST_Expand expands the bounding box an x,y,z unit amount.
Units are in the units of the spatial reference system in use denoted by the SRID
Pre 1.3, ST_Expand was used in conjunction with distance to do indexable queries. Something of the form |
Availability: 1.5.0 behavior changed to output double precision instead of float4 coordinates. |
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
Examples below use US National Atlas Equal Area (SRID=2163) which is a meter projection |
--10 meter expanded box around bbox of a linestring SELECT CAST(ST_Expand(ST_GeomFromText('LINESTRING(2312980 110676,2312923 110701,2312892 110714)', 2163),10) As box2d); st_expand ------------------------------------ BOX(2312882 110666,2312990 110724) --10 meter expanded 3d box of a 3d box SELECT ST_Expand(CAST('BOX3D(778783 2951741 1,794875 2970042.61545891 10)' As box3d),10) st_expand ----------------------------------------------------- BOX3D(778773 2951731 -9,794885 2970052.61545891 20) --10 meter geometry astext rep of a expand box around a point geometry SELECT ST_AsEWKT(ST_Expand(ST_GeomFromEWKT('SRID=2163;POINT(2312980 110676)'),10)); st_asewkt ------------------------------------------------------------------------------------------------- SRID=2163;POLYGON((2312970 110666,2312970 110686,2312990 110686,2312990 110666,2312970 110666))
ST_Extent — an aggregate function that returns the bounding box that bounds rows of geometries.
box2d ST_Extent(
geometry set geomfield)
;
ST_Extent returns a bounding box that encloses a set of geometries. The ST_Extent function is an "aggregate" function in the terminology of SQL. That means that it operates on lists of data, in the same way the SUM() and AVG() functions do.
Since it returns a bounding box, the spatial Units are in the units of the spatial reference system in use denoted by the SRID
ST_Extent is similar in concept to Oracle Spatial/Locator's SDO_AGGR_MBR
Since ST_Extent returns a bounding box, the SRID meta-data is lost. Use ST_SetSRID to force it back into a geometry with SRID meta data. The coordinates are in the units of the spatial ref of the orginal geometries. |
ST_Extent will return boxes with only an x and y component even with (x,y,z) coordinate geometries. To maintain x,y,z use ST_3DExtent instead. |
Availability: 1.4.0 |
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
Examples below use Massachusetts State Plane ft (SRID=2249) |
SELECT ST_Extent(the_geom) as bextent FROM sometable; st_bextent ------------------------------------ BOX(739651.875 2908247.25,794875.8125 2970042.75) -- restituisce l'estensione di ciascuna categoria di geometrie SELECT ST_Extent(the_geom) as bextent FROM sometable GROUP BY category ORDER BY category; bextent | name ----------------------------------------------------+---------------- BOX(778783.5625 2951741.25,794875.8125 2970042.75) | A BOX(751315.8125 2919164.75,765202.6875 2935417.25) | B BOX(739651.875 2917394.75,756688.375 2935866) | C --riporta il risultato a una geometria -- e visualizza la rappresentazione come testo di quella geometria SELECT ST_SetSRID(ST_Extent(the_geom),2249) as bextent FROM sometable; bextent -------------------------------------------------------------------------------- SRID=2249;POLYGON((739651.875 2908247.25,739651.875 2970042.75,794875.8125 2970042.75, 794875.8125 2908247.25,739651.875 2908247.25))
ST_3DExtent — an aggregate function that returns the box3D bounding box that bounds rows of geometries.
box3d ST_3DExtent(
geometry set geomfield)
;
ST_3DExtent returns a box3d (includes Z coordinate) bounding box that encloses a set of geometries. The ST_3DExtent function is an "aggregate" function in the terminology of SQL. That means that it operates on lists of data, in the same way the SUM() and AVG() functions do.
Since it returns a bounding box, the spatial Units are in the units of the spatial reference system in use denoted by the SRID
Since ST_3DExtent returns a bounding box, the SRID meta-data is lost. Use ST_SetSRID to force it back into a geometry with SRID meta data. The coordinates are in the units of the spatial ref of the orginal geometries. |
Enhanced: 2.0.0 support for Polyhedral surfaces, Triangles and TIN was introduced.
Modificato nella versione 2.0.0. Nelle versioni precedenti era chiamato ST_Extent3D
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
SELECT ST_3DExtent(foo.the_geom) As b3extent FROM (SELECT ST_MakePoint(x,y,z) As the_geom FROM generate_series(1,3) As x CROSS JOIN generate_series(1,2) As y CROSS JOIN generate_series(0,2) As Z) As foo; b3extent -------------------- BOX3D(1 1 0,3 2 2) --Ottieni l'estensione di varie stringhe circolari con elevazione SELECT ST_3DExtent(foo.the_geom) As b3extent FROM (SELECT ST_Translate(ST_Force_3DZ(ST_LineToCurve(ST_Buffer(ST_MakePoint(x,y),1))),0,0,z) As the_geom FROM generate_series(1,3) As x CROSS JOIN generate_series(1,2) As y CROSS JOIN generate_series(0,2) As Z) As foo; b3extent -------------------- BOX3D(1 0 0,4 2 2)
Find_SRID — The syntax is find_srid(a_db_schema, a_table, a_column) and the function returns the integer SRID of the specified column by searching through the GEOMETRY_COLUMNS table.
integer Find_SRID(
varchar a_schema_name, varchar a_table_name, varchar a_geomfield_name)
;
The syntax is find_srid(<db/schema>, <table>, <column>) and the function returns the integer SRID of the specified column by searching through the GEOMETRY_COLUMNS table. If the geometry column has not been properly added with the AddGeometryColumns() function, this function will not work either.
ST_MemSize — Restituisce lo spazio occupato da una geometria (in byte).
integer ST_MemSize(
geometry geomA)
;
Restituisce lo spazio occupato da una geometria (in byte).
This is a nice compliment to PostgreSQL built in functions pg_column_size, pg_size_pretty, pg_relation_size, pg_total_relation_size.
pg_relation_size which gives the byte size of a table may return byte size lower than ST_MemSize. This is because pg_relation_size does not add toasted table contribution and large geometries are stored in TOAST tables. pg_total_relation_size - includes, the table, the toasted tables, and the indexes. pg_column_size returns how much space a geometry would take in a column considering compression, so may be lower than ST_MemSize |
This function supports 3d and will not drop the z-index.
This method supports Circular Strings and Curves
This function supports Polyhedral surfaces.
This function supports Triangles and Triangulated Irregular Network Surfaces (TIN).
Changed: 2.2.0 name changed to ST_MemSize to follow naming convention. In prior versions this function was called ST_Mem_Size, old name deprecated though still available.
--Return how much byte space Boston takes up in our Mass data set SELECT pg_size_pretty(SUM(ST_MemSize(the_geom))) as totgeomsum, pg_size_pretty(SUM(CASE WHEN town = 'BOSTON' THEN ST_MemSize(the_geom) ELSE 0 END)) As bossum, CAST(SUM(CASE WHEN town = 'BOSTON' THEN ST_MemSize(the_geom) ELSE 0 END)*1.00 / SUM(ST_MemSize(the_geom))*100 As numeric(10,2)) As perbos FROM towns; totgeomsum bossum perbos ---------- ------ ------ 1522 kB 30 kB 1.99 SELECT ST_MemSize(ST_GeomFromText('CIRCULARSTRING(220268 150415,220227 150505,220227 150406)')); --- 73 --What percentage of our table is taken up by just the geometry SELECT pg_total_relation_size('public.neighborhoods') As fulltable_size, sum(ST_MemSize(the_geom)) As geomsize, sum(ST_MemSize(the_geom))*1.00/pg_total_relation_size('public.neighborhoods')*100 As pergeom FROM neighborhoods; fulltable_size geomsize pergeom ------------------------------------------------ 262144 96238 36.71188354492187500000
ST_PointInsideCircle — Is the point geometry insert circle defined by center_x, center_y, radius
boolean ST_PointInsideCircle(
geometry a_point, float center_x, float center_y, float radius)
;
The syntax for this functions is ST_PointInsideCircle(<geometry>,<circle_center_x>,<circle_center_y>,<radius>). Returns the true if the geometry is a point and is inside the circle. Returns false otherwise.
Come suggerito dal nome, funziona solo per i punti |
Availability: 1.2
Changed: 2.2.0 In prior versions this used to be called ST_Point_Inside_Circle
These functions are rarely used functions that should only be used if your data is corrupted in someway. They are used for troubleshooting corruption and also fixing things that should under normal circumstances, never happen.
PostGIS_AddBBox — Add bounding box to the geometry.
geometry PostGIS_AddBBox(
geometry geomA)
;
Add bounding box to the geometry. This would make bounding box based queries faster, but will increase the size of the geometry.
Bounding boxes are automatically added to geometries so in general this is not needed unless the generated bounding box somehow becomes corrupted or you have an old install that is lacking bounding boxes. Then you need to drop the old and readd. |
This method supports Circular Strings and Curves
PostGIS_DropBBox — Drop the bounding box cache from the geometry.
geometry PostGIS_DropBBox(
geometry geomA)
;
Drop the bounding box cache from the geometry. This reduces geometry size, but makes bounding-box based queries slower. It is also used to drop a corrupt bounding box. A tale-tell sign of a corrupt cached bounding box is when your ST_Intersects and other relation queries leave out geometries that rightfully should return true.
Bounding boxes are automatically added to geometries and improve speed of queries so in general this is not needed unless the generated bounding box somehow becomes corrupted or you have an old install that is lacking bounding boxes. Then you need to drop the old and readd. This kind of corruption has been observed in 8.3-8.3.6 series whereby cached bboxes were not always recalculated when a geometry changed and upgrading to a newer version without a dump reload will not correct already corrupted boxes. So one can manually correct using below and readd the bbox or do a dump reload. |
This method supports Circular Strings and Curves
--This example drops bounding boxes where the cached box is not correct --The force to ST_AsBinary before applying Box2D forces a recalculation of the box, and Box2D applied to the table geometry always -- returns the cached bounding box. UPDATE sometable SET the_geom = PostGIS_DropBBox(the_geom) WHERE Not (Box2D(ST_AsBinary(the_geom)) = Box2D(the_geom)); UPDATE sometable SET the_geom = PostGIS_AddBBox(the_geom) WHERE Not PostGIS_HasBBOX(the_geom);
PostGIS_HasBBox — Returns TRUE if the bbox of this geometry is cached, FALSE otherwise.
boolean PostGIS_HasBBox(
geometry geomA)
;
Returns TRUE if the bbox of this geometry is cached, FALSE otherwise. Use PostGIS_AddBBox and PostGIS_DropBBox to control caching.
This method supports Circular Strings and Curves
Table of Contents
Le funzioni riportate sotto sono quelle che un utente di PostGIS Raster si troverà a utilizzare più frequentemente e che sono disponibili in PostGIS Raster. Esistono altre funzioni che sono di supporto agli oggetti raster e che non servono a un utilizzo generale.
Il tipo raster
è un nuovo tipo di dati PostGIS per archiviare e analizzare dati raster.
For loading rasters from raster files please refer to Section 5.1, “Caricare e creare raster”
For the examples in this reference we will be using a raster table of dummy rasters - Formed with the following code
CREATE TABLE dummy_rast(rid integer, rast raster); INSERT INTO dummy_rast(rid, rast) VALUES (1, ('01' -- little endian (uint8 ndr) || '0000' -- version (uint16 0) || '0000' -- nBands (uint16 0) || '0000000000000040' -- scaleX (float64 2) || '0000000000000840' -- scaleY (float64 3) || '000000000000E03F' -- ipX (float64 0.5) || '000000000000E03F' -- ipY (float64 0.5) || '0000000000000000' -- skewX (float64 0) || '0000000000000000' -- skewY (float64 0) || '00000000' -- SRID (int32 0) || '0A00' -- width (uint16 10) || '1400' -- height (uint16 20) )::raster ), -- Raster: 5 x 5 pixels, 3 bands, PT_8BUI pixel type, NODATA = 0 (2, ('01000003009A9999999999A93F9A9999999999A9BF000000E02B274A' || '41000000007719564100000000000000000000000000000000FFFFFFFF050005000400FDFEFDFEFEFDFEFEFDF9FAFEF' || 'EFCF9FBFDFEFEFDFCFAFEFEFE04004E627AADD16076B4F9FE6370A9F5FE59637AB0E54F58617087040046566487A1506CA2E3FA5A6CAFFBFE4D566DA4CB3E454C5665')::raster);
geomval — A spatial datatype with two fields - geom (holding a geometry object) and val (holding a double precision pixel value from a raster band).
geomval is a compound data type consisting of a geometry object referenced by the .geom field and val, a double precision value that represents the pixel value at a particular geometric location in a raster band. It is used by the ST_DumpAsPolygon and Raster intersection family of functions as an output type to explode a raster band into geometry polygons.
addbandarg — A composite type used as input into the ST_AddBand function defining the attributes and initial value of the new band.
A composite type used as input into the ST_AddBand function defining the attributes and initial value of the new band.
index
integer 1-based value indicating the position where the new band will be added amongst the raster's bands. If NULL, the new band will be added at the end of the raster's bands.
pixeltype
text Pixel type of the new band. One of defined pixel types as described in ST_BandPixelType.
initialvalue
double precision Initial value that all pixels of new band will be set to.
nodataval
double precision NODATA value of the new band. If NULL, the new band will not have a NODATA value assigned.
rastbandarg — A composite type for use when needing to express a raster and a band index of that raster.
raster — raster spatial data type.
raster is a spatial data type used to represent raster data such as those imported from jpegs, tiffs, pngs, digital elevation models. Each raster has 1 or more bands each having a set of pixel values. Rasters can be georeferenced.
Requires PostGIS be compiled with GDAL support. Currently rasters can be implicitly converted to geometry type, but the conversion returns the ST_ConvexHull of the raster. This auto casting may be removed in the near future so don't rely on it. |
reclassarg — A composite type used as input into the ST_Reclass function defining the behavior of reclassification.
A composite type used as input into the ST_Reclass function defining the behavior of reclassification.
nband
integerThe band number of band to reclassify.
reclassexpr
textrange expression consisting of comma delimited range:map_range mappings. : to define mapping that defines how to map old band values to new band values. ( means >, ) means less than, ] < or equal, [ means > or equal
1. [a-b] = a <= x <= b 2. (a-b] = a < x <= b 3. [a-b) = a <= x < b 4. (a-b) = a < x < b
( notation is optional so a-b means the same as (a-b)
pixeltype
textOne of defined pixel types as described in ST_BandPixelType
nodataval
double precisionValue to treat as no data. For image outputs that support transparency, these will be blank.
SELECT ROW(2, '0-100:1-10, 101-500:11-150,501 - 10000: 151-254', '8BUI', 255)::reclassarg;
summarystats — A composite type returned by the ST_SummaryStats and ST_SummaryStatsAgg functions.
A composite type returned by the ST_SummaryStats and ST_SummaryStatsAgg functions.
count
integer Number of pixels counted for the summary statistics.
sum
double precision Sum of all counted pixel values.
mean
double precision Arithmetic mean of all counted pixel values.
stddev
double precision Standard deviation of all counted pixel values.
min
double precision Minimum value of counted pixel values.
max
double precision Maximum value of counted pixel values.
unionarg — A composite type used as input into the ST_Union function defining the bands to be processed and behavior of the UNION operation.
A composite type used as input into the ST_Union function defining the bands to be processed and behavior of the UNION operation.
nband
integer 1-based value indicating the band of each input raster to be processed.
uniontype
text Type of UNION operation. One of defined types as described in ST_Union.
AddRasterConstraints — Adds raster constraints to a loaded raster table for a specific column that constrains spatial ref, scaling, blocksize, alignment, bands, band type and a flag to denote if raster column is regularly blocked. The table must be loaded with data for the constraints to be inferred. Returns true of the constraint setting was accomplished and if issues a notice.
boolean AddRasterConstraints(
name rasttable, name rastcolumn, boolean srid, boolean scale_x, boolean scale_y, boolean blocksize_x, boolean blocksize_y, boolean same_alignment, boolean regular_blocking, boolean num_bands=true , boolean pixel_types=true , boolean nodata_values=true , boolean out_db=true , boolean extent=true )
;
boolean AddRasterConstraints(
name rasttable, name rastcolumn, text[] VARIADIC constraints)
;
boolean AddRasterConstraints(
name rastschema, name rasttable, name rastcolumn, text[] VARIADIC constraints)
;
boolean AddRasterConstraints(
name rastschema, name rasttable, name rastcolumn, boolean srid=true, boolean scale_x=true, boolean scale_y=true, boolean blocksize_x=true, boolean blocksize_y=true, boolean same_alignment=true, boolean regular_blocking=false, boolean num_bands=true, boolean pixel_types=true, boolean nodata_values=true , boolean out_db=true , boolean extent=true )
;
Generates constraints on a raster column that are used to display information in the raster_columns
raster catalog. The rastschema
is the name of the table schema the table resides in. The srid
must be an integer value reference to an entry in the SPATIAL_REF_SYS table.
raster2pgsql
loader uses this function to register raster tables
Valid constraint names to pass in: refer to Section 5.2.1, “Catalogo delle colonne raster” for more details.
blocksize
sets both X and Y blocksize
blocksize_x
sets X tile (width in pixels of each tile)
blocksize_y
sets Y tile (height in pixels of each tile)
extent
computes extent of whole table and applys constraint all rasters must be within that extent
num_bands
number of bands
pixel_types
reads array of pixel types for each band ensure all band n have same pixel type
regular_blocking
sets spatially unique (no two rasters can be spatially the same) and coverage tile (raster is aligned to a coverage) constraints
same_alignment
ensures they all have same alignment meaning any two tiles you compare will return true for. Refer to ST_SameAlignment
srid
ensures all have same srid
More -- any listed as inputs into the above functions
This function infers the constraints from the data already present in the table. As such for it to work, you must create the raster column first and then load it with data. |
If you need to load more data in your tables after you have already applied constraints, you may want to run the DropRasterConstraints if the extent of your data has changed. |
Disponibilità: 2.0
CREATE TABLE myrasters(rid SERIAL primary key, rast raster); INSERT INTO myrasters(rast) SELECT ST_AddBand(ST_MakeEmptyRaster(1000, 1000, 0.3, -0.3, 2, 2, 0, 0,4326), 1, '8BSI'::text, -129, NULL); SELECT AddRasterConstraints('myrasters'::name, 'rast'::name); -- verify if registered correctly in the raster_columns view -- SELECT srid, scale_x, scale_y, blocksize_x, blocksize_y, num_bands, pixel_types, nodata_values FROM raster_columns WHERE r_table_name = 'myrasters'; srid | scale_x | scale_y | blocksize_x | blocksize_y | num_bands | pixel_types| nodata_values ------+---------+---------+-------------+-------------+-----------+-------------+--------------- 4326 | 2 | 2 | 1000 | 1000 | 1 | {8BSI} | {0}
CREATE TABLE public.myrasters2(rid SERIAL primary key, rast raster); INSERT INTO myrasters2(rast) SELECT ST_AddBand(ST_MakeEmptyRaster(1000, 1000, 0.3, -0.3, 2, 2, 0, 0,4326), 1, '8BSI'::text, -129, NULL); SELECT AddRasterConstraints('public'::name, 'myrasters2'::name, 'rast'::name,'regular_blocking', 'blocksize'); -- get notice-- NOTICE: Adding regular blocking constraint NOTICE: Adding blocksize-X constraint NOTICE: Adding blocksize-Y constraint
DropRasterConstraints — Drops PostGIS raster constraints that refer to a raster table column. Useful if you need to reload data or update your raster column data.
boolean DropRasterConstraints(
name rasttable, name rastcolumn, boolean srid, boolean scale_x, boolean scale_y, boolean blocksize_x, boolean blocksize_y, boolean same_alignment, boolean regular_blocking, boolean num_bands=true, boolean pixel_types=true, boolean nodata_values=true, boolean out_db=true , boolean extent=true)
;
boolean DropRasterConstraints(
name rastschema, name rasttable, name rastcolumn, boolean srid=true, boolean scale_x=true, boolean scale_y=true, boolean blocksize_x=true, boolean blocksize_y=true, boolean same_alignment=true, boolean regular_blocking=false, boolean num_bands=true, boolean pixel_types=true, boolean nodata_values=true, boolean out_db=true , boolean extent=true)
;
boolean DropRasterConstraints(
name rastschema, name rasttable, name rastcolumn, text[] constraints)
;
Drops PostGIS raster constraints that refer to a raster table column that were added by AddRasterConstraints. Useful if you need to load more data or update your raster column data. You do not need to do this if you want to get rid of a raster table or a raster column.
To drop a raster table use the standard
DROP TABLE mytable
To drop just a raster column and leave the rest of the table, use standard SQL
ALTER TABLE mytable DROP COLUMN rast
the table will disappear from the raster_columns
catalog if the column or table is dropped. However if only the constraints are dropped, the raster column will still be listed in the raster_columns
catalog, but there will be no other information about it aside from the column name and table.
Disponibilità: 2.0
SELECT DropRasterConstraints ('myrasters','rast'); ----RESULT output --- t -- verify change in raster_columns -- SELECT srid, scale_x, scale_y, blocksize_x, blocksize_y, num_bands, pixel_types, nodata_values FROM raster_columns WHERE r_table_name = 'myrasters'; srid | scale_x | scale_y | blocksize_x | blocksize_y | num_bands | pixel_types| nodata_values ------+---------+---------+-------------+-------------+-----------+-------------+--------------- 0 | | | | | | |
AddOverviewConstraints — Tag a raster column as being an overview of another.
boolean AddOverviewConstraints(
name ovschema, name ovtable, name ovcolumn, name refschema, name reftable, name refcolumn, int ovfactor)
;
boolean AddOverviewConstraints(
name ovtable, name ovcolumn, name reftable, name refcolumn, int ovfactor)
;
Adds constraints on a raster column that are used to display information in the raster_overviews
raster catalog.
The ovfactor
parameter represents the scale multiplier in the overview column: higher overview factors have lower resolution.
When the ovschema
and refschema
parameters are omitted, the first table found scanning the search_path
will be used.
Disponibilità: 2.0
CREATE TABLE res1 AS SELECT ST_AddBand( ST_MakeEmptyRaster(1000, 1000, 0, 0, 2), 1, '8BSI'::text, -129, NULL ) r1; CREATE TABLE res2 AS SELECT ST_AddBand( ST_MakeEmptyRaster(500, 500, 0, 0, 4), 1, '8BSI'::text, -129, NULL ) r2; SELECT AddOverviewConstraints('res2', 'r2', 'res1', 'r1', 2); -- verify if registered correctly in the raster_overviews view -- SELECT o_table_name ot, o_raster_column oc, r_table_name rt, r_raster_column rc, overview_factor f FROM raster_overviews WHERE o_table_name = 'res2'; ot | oc | rt | rc | f ------+----+------+----+--- res2 | r2 | res1 | r1 | 2 (1 row)
DropOverviewConstraints — Untag a raster column from being an overview of another.
boolean DropOverviewConstraints(
name ovschema, name ovtable, name ovcolumn)
;
boolean DropOverviewConstraints(
name ovtable, name ovcolumn)
;
PostGIS_GDAL_Version — Reports the version of the GDAL library in use by PostGIS.
text PostGIS_GDAL_Version(
)
;
Reports the version of the GDAL library in use by PostGIS. Will also check and report if GDAL can find its data files.
PostGIS_Raster_Lib_Build_Date — Reports full raster library build date.
text PostGIS_Raster_Lib_Build_Date(
)
;
PostGIS_Raster_Lib_Version — Reports full raster version and build configuration infos.
text PostGIS_Raster_Lib_Version(
)
;
ST_GDALDrivers — Returns a list of raster formats supported by your lib gdal. These are the formats you can output your raster using ST_AsGDALRaster.
setof record ST_GDALDrivers(
integer OUT idx, text OUT short_name, text OUT long_name, text OUT create_options)
;
Returns a list of raster formats short_name,long_name and creator options of each format supported by your lib gdal. Use the short_name as input in the format
parameter of ST_AsGDALRaster. Options vary depending on what drivers your libgdal was compiled with. create_options
returns an xml formatted set of CreationOptionList/Option consisting of name and optional type
, description
and set of VALUE
for each creator option for the specific driver.
Changed: 2.0.6, 2.1.3 - by default no drivers are enabled, unless GUC or Environment variable gdal_enabled_drivers is set.
Availability: 2.0.0 - requires GDAL >= 1.6.0.
SET postgis.gdal_enabled_drivers = 'ENABLE_ALL'; SELECT short_name, long_name FROM st_gdaldrivers() ORDER BY short_name; short_name | long_name -----------------+-------------------------------------- AAIGrid | Arc/Info ASCII Grid ARG | Azavea Raster Grid format DTED | DTED Elevation Raster EHdr | ESRI .hdr Labelled FIT | FIT Image GIF | Graphics Interchange Format (.gif) GPKG | GeoPackage GS7BG | Golden Software 7 Binary Grid (.grd) GSAG | Golden Software ASCII Grid (.grd) GSBG | Golden Software Binary Grid (.grd) GTiff | GeoTIFF HF2 | HF2/HFZ heightfield raster HFA | Erdas Imagine Images (.img) ILWIS | ILWIS Raster Map INGR | Intergraph Raster JPEG | JPEG JFIF KMLSUPEROVERLAY | Kml Super Overlay LCP | FARSITE v.4 Landscape File (.lcp) MFF | Vexcel MFF Raster NITF | National Imagery Transmission Format PNG | Portable Network Graphics R | R Object Data Store RST | Idrisi Raster A.1 SAGA | SAGA GIS Binary Grid (.sdat) SRTMHGT | SRTMHGT File Format USGSDEM | USGS Optional ASCII DEM (and CDED) VRT | Virtual Raster WMS | OGC Web Map Service XPM | X11 PixMap Format XYZ | ASCII Gridded XYZ ZMap | ZMap Plus Grid (31 rows)
-- Output the create options XML column of JPEG as a table -- -- Note you can use these creator options in ST_AsGDALRaster options argument SELECT (xpath('@name', g.opt))[1]::text As oname, (xpath('@type', g.opt))[1]::text As otype, (xpath('@description', g.opt))[1]::text As descrip FROM (SELECT unnest(xpath('/CreationOptionList/Option', create_options::xml)) As opt FROM st_gdaldrivers() WHERE short_name = 'JPEG') As g; oname | otype | descrip --------------------+---------+------------------------------------------------- PROGRESSIVE | boolean | whether to generate a progressive JPEG QUALITY | int | good=100, bad=0, default=75 WORLDFILE | boolean | whether to geneate a worldfile INTERNAL_MASK | boolean | whether to generate a validity mask COMMENT | string | Comment SOURCE_ICC_PROFILE | string | ICC profile encoded in Base64 EXIF_THUMBNAIL | boolean | whether to generate an EXIF thumbnail(overview). By default its max dimension will be 128 THUMBNAIL_WIDTH | int | Forced thumbnail width THUMBNAIL_HEIGHT | int | Forced thumbnail height (9 rows)
-- raw xml output for creator options for GeoTiff -- SELECT create_options FROM st_gdaldrivers() WHERE short_name = 'GTiff'; <CreationOptionList> <Option name="COMPRESS" type="string-select"> <Value>NONE</Value> <Value>LZW</Value> <Value>PACKBITS</Value> <Value>JPEG</Value> <Value>CCITTRLE</Value> <Value>CCITTFAX3</Value> <Value>CCITTFAX4</Value> <Value>DEFLATE</Value> </Option> <Option name="PREDICTOR" type="int" description="Predictor Type"/> <Option name="JPEG_QUALITY" type="int" description="JPEG quality 1-100" default="75"/> <Option name="ZLEVEL" type="int" description="DEFLATE compression level 1-9" default="6"/> <Option name="NBITS" type="int" description="BITS for sub-byte files (1-7), sub-uint16 (9-15), sub-uint32 (17-31)"/> <Option name="INTERLEAVE" type="string-select" default="PIXEL"> <Value>BAND</Value> <Value>PIXEL</Value> </Option> <Option name="TILED" type="boolean" description="Switch to tiled format"/> <Option name="TFW" type="boolean" description="Write out world file"/> <Option name="RPB" type="boolean" description="Write out .RPB (RPC) file"/> <Option name="BLOCKXSIZE" type="int" description="Tile Width"/> <Option name="BLOCKYSIZE" type="int" description="Tile/Strip Height"/> <Option name="PHOTOMETRIC" type="string-select"> <Value>MINISBLACK</Value> <Value>MINISWHITE</Value> <Value>PALETTE</Value> <Value>RGB</Value> <Value>CMYK</Value> <Value>YCBCR</Value> <Value>CIELAB</Value> <Value>ICCLAB</Value> <Value>ITULAB</Value> </Option> <Option name="SPARSE_OK" type="boolean" description="Can newly created files have missing blocks?" default="FALSE"/> <Option name="ALPHA" type="boolean" description="Mark first extrasample as being alpha"/> <Option name="PROFILE" type="string-select" default="GDALGeoTIFF"> <Value>GDALGeoTIFF</Value> <Value>GeoTIFF</Value> <Value>BASELINE</Value> </Option> <Option name="PIXELTYPE" type="string-select"> <Value>DEFAULT</Value> <Value>SIGNEDBYTE</Value> </Option> <Option name="BIGTIFF" type="string-select" description="Force creation of BigTIFF file"> <Value>YES</Value> <Value>NO</Value> <Value>IF_NEEDED</Value> <Value>IF_SAFER</Value> </Option> <Option name="ENDIANNESS" type="string-select" default="NATIVE" description="Force endianness of created file. For DEBUG purpose mostly"> <Value>NATIVE</Value> <Value>INVERTED</Value> <Value>LITTLE</Value> <Value>BIG</Value> </Option> <Option name="COPY_SRC_OVERVIEWS" type="boolean" default="NO" description="Force copy of overviews of source dataset (CreateCopy())"/> </CreationOptionList> -- Output the create options XML column for GTiff as a table -- SELECT (xpath('@name', g.opt))[1]::text As oname, (xpath('@type', g.opt))[1]::text As otype, (xpath('@description', g.opt))[1]::text As descrip, array_to_string(xpath('Value/text()', g.opt),', ') As vals FROM (SELECT unnest(xpath('/CreationOptionList/Option', create_options::xml)) As opt FROM st_gdaldrivers() WHERE short_name = 'GTiff') As g; oname | otype | descrip | vals --------------------+---------------+----------------------------------------------------------------------+--------------------------------------------------------------------------- COMPRESS | string-select | | NONE, LZW, PACKBITS, JPEG, CCITTRLE, CCITTFAX3, CCITTFAX4, DEFLATE PREDICTOR | int | Predictor Type | JPEG_QUALITY | int | JPEG quality 1-100 | ZLEVEL | int | DEFLATE compression level 1-9 | NBITS | int | BITS for sub-byte files (1-7), sub-uint16 (9-15), sub-uint32 (17-31) | INTERLEAVE | string-select | | BAND, PIXEL TILED | boolean | Switch to tiled format | TFW | boolean | Write out world file | RPB | boolean | Write out .RPB (RPC) file | BLOCKXSIZE | int | Tile Width | BLOCKYSIZE | int | Tile/Strip Height | PHOTOMETRIC | string-select | | MINISBLACK, MINISWHITE, PALETTE, RGB, CMYK, YCBCR, CIELAB, ICCLAB, ITULAB SPARSE_OK | boolean | Can newly created files have missing blocks? | ALPHA | boolean | Mark first extrasample as being alpha | PROFILE | string-select | | GDALGeoTIFF, GeoTIFF, BASELINE PIXELTYPE | string-select | | DEFAULT, SIGNEDBYTE BIGTIFF | string-select | Force creation of BigTIFF file | YES, NO, IF_NEEDED, IF_SAFER ENDIANNESS | string-select | Force endianness of created file. For DEBUG purpose mostly | NATIVE, INVERTED, LITTLE, BIG COPY_SRC_OVERVIEWS | boolean | Force copy of overviews of source dataset (CreateCopy()) | (19 rows)
UpdateRasterSRID — Change the SRID of all rasters in the user-specified column and table.
raster UpdateRasterSRID(
name schema_name, name table_name, name column_name, integer new_srid)
;
raster UpdateRasterSRID(
name table_name, name column_name, integer new_srid)
;
Change the SRID of all rasters in the user-specified column and table. The function will drop all appropriate column constraints (extent, alignment and SRID) before changing the SRID of the specified column's rasters.
The data (band pixel values) of the rasters are not touched by this function. Only the raster's metadata is changed. |
Availability: 2.1.0
ST_CreateOverview — Create an reduced resolution version of a given raster coverage.
regclass ST_CreateOverview(
regclass tab, name col, int factor, text algo='NearestNeighbor')
;
Create an overview table with resampled tiles from the source table. Output tiles will have the same size of input tiles and cover the same spatial extent with a lower resolution (pixel size will be 1/factor
of the original in both directions).
The overview table will be made available in the raster_overviews
catalog and will have raster constraints enforced.
Algorithm options are: 'NearestNeighbor', 'Bilinear', 'Cubic', 'CubicSpline', and 'Lanczos'. Refer to: GDAL Warp resampling methods for more details.
Availability: 2.2.0
ST_AddBand — Returns a raster with the new band(s) of given type added with given initial value in the given index location. If no index is specified, the band is added to the end.
(1) raster ST_AddBand(
raster rast, addbandarg[] addbandargset)
;
(2) raster ST_AddBand(
raster rast, integer index, text pixeltype, double precision initialvalue=0, double precision nodataval=NULL)
;
(3) raster ST_AddBand(
raster rast, text pixeltype, double precision initialvalue=0, double precision nodataval=NULL)
;
(4) raster ST_AddBand(
raster torast, raster fromrast, integer fromband=1, integer torastindex=at_end)
;
(5) raster ST_AddBand(
raster torast, raster[] fromrasts, integer fromband=1, integer torastindex=at_end)
;
(6) raster ST_AddBand(
raster rast, integer index, text outdbfile, integer[] outdbindex, double precision nodataval=NULL)
;
(7) raster ST_AddBand(
raster rast, text outdbfile, integer[] outdbindex, integer index=at_end, double precision nodataval=NULL)
;
Returns a raster with a new band added in given position (index), of given type, of given initial value, and of given nodata value. If no index is specified, the band is added to the end. If no fromband
is specified, band 1 is assumed. Pixel type is a string representation of one of the pixel types specified in ST_BandPixelType. If an existing index is specified all subsequent bands
>= that index are incremented by 1. If an initial value greater than the max of the pixel type is specified, then the initial value is set to the highest value allowed by the pixel type.
For the variant that takes an array of addbandarg (Variant 1), a specific addbandarg's index value is relative to the raster at the time when the band described by that addbandarg is being added to the raster. See the Multiple New Bands example below.
For the variant that takes an array of rasters (Variant 5), if torast
is NULL then the fromband
band of each raster in the array is accumulated into a new raster.
For the variants that take outdbfile
(Variants 6 and 7), the value must include the full path to the raster file. The file must also be accessible to the postgres server process.
Enhanced: 2.1.0 support for addbandarg added.
Enhanced: 2.1.0 support for new out-db bands added.
-- Add another band of type 8 bit unsigned integer with pixels initialized to 200 UPDATE dummy_rast SET rast = ST_AddBand(rast,'8BUI'::text,200) WHERE rid = 1;
-- Create an empty raster 100x100 units, with upper left right at 0, add 2 bands (band 1 is 0/1 boolean bit switch, band2 allows values 0-15) -- uses addbandargs INSERT INTO dummy_rast(rid,rast) VALUES(10, ST_AddBand(ST_MakeEmptyRaster(100, 100, 0, 0, 1, -1, 0, 0, 0), ARRAY[ ROW(1, '1BB'::text, 0, NULL), ROW(2, '4BUI'::text, 0, NULL) ]::addbandarg[] ) ); -- output meta data of raster bands to verify all is right -- SELECT (bmd).* FROM (SELECT ST_BandMetaData(rast,generate_series(1,2)) As bmd FROM dummy_rast WHERE rid = 10) AS foo; --result -- pixeltype | nodatavalue | isoutdb | path -----------+----------------+-------------+---------+------ 1BB | | f | 4BUI | | f | -- output meta data of raster - SELECT (rmd).width, (rmd).height, (rmd).numbands FROM (SELECT ST_MetaData(rast) As rmd FROM dummy_rast WHERE rid = 10) AS foo; -- result -- upperleftx | upperlefty | width | height | scalex | scaley | skewx | skewy | srid | numbands ------------+------------+-------+--------+------------+------------+-------+-------+------+---------- 0 | 0 | 100 | 100 | 1 | -1 | 0 | 0 | 0 | 2
SELECT * FROM ST_BandMetadata( ST_AddBand( ST_MakeEmptyRaster(10, 10, 0, 0, 1, -1, 0, 0, 0), ARRAY[ ROW(NULL, '8BUI', 255, 0), ROW(NULL, '16BUI', 1, 2), ROW(2, '32BUI', 100, 12), ROW(2, '32BF', 3.14, -1) ]::addbandarg[] ), ARRAY[]::integer[] ); bandnum | pixeltype | nodatavalue | isoutdb | path ---------+-----------+-------------+---------+------ 1 | 8BUI | 0 | f | 2 | 32BF | -1 | f | 3 | 32BUI | 12 | f | 4 | 16BUI | 2 | f |
-- Aggregate the 1st band of a table of like rasters into a single raster -- with as many bands as there are test_types and as many rows (new rasters) as there are mice -- NOTE: The ORDER BY test_type is only supported in PostgreSQL 9.0+ -- for 8.4 and below it usually works to order your data in a subselect (but not guaranteed) -- The resulting raster will have a band for each test_type alphabetical by test_type -- For mouse lovers: No mice were harmed in this exercise SELECT mouse, ST_AddBand(NULL, array_agg(rast ORDER BY test_type), 1) As rast FROM mice_studies GROUP BY mouse;
SELECT * FROM ST_BandMetadata( ST_AddBand( ST_MakeEmptyRaster(10, 10, 0, 0, 1, -1, 0, 0, 0), '/home/raster/mytestraster.tif'::text, NULL::int[] ), ARRAY[]::integer[] ); bandnum | pixeltype | nodatavalue | isoutdb | path ---------+-----------+-------------+---------+------ 1 | 8BUI | | t | /home/raster/mytestraster.tif 2 | 8BUI | | t | /home/raster/mytestraster.tif 3 | 8BUI | | t | /home/raster/mytestraster.tif
ST_BandMetaData, ST_BandPixelType, ST_MakeEmptyRaster, ST_MetaData, ST_NumBands, ST_Reclass
ST_AsRaster — Converts a PostGIS geometry to a PostGIS raster.
raster ST_AsRaster(
geometry geom, raster ref, text pixeltype, double precision value=1, double precision nodataval=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, raster ref, text[] pixeltype=ARRAY['8BUI'], double precision[] value=ARRAY[1], double precision[] nodataval=ARRAY[0], boolean touched=false)
;
raster ST_AsRaster(
geometry geom, double precision scalex, double precision scaley, double precision gridx, double precision gridy, text pixeltype, double precision value=1, double precision nodataval=0, double precision skewx=0, double precision skewy=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, double precision scalex, double precision scaley, double precision gridx=NULL, double precision gridy=NULL, text[] pixeltype=ARRAY['8BUI'], double precision[] value=ARRAY[1], double precision[] nodataval=ARRAY[0], double precision skewx=0, double precision skewy=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, double precision scalex, double precision scaley, text pixeltype, double precision value=1, double precision nodataval=0, double precision upperleftx=NULL, double precision upperlefty=NULL, double precision skewx=0, double precision skewy=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, double precision scalex, double precision scaley, text[] pixeltype, double precision[] value=ARRAY[1], double precision[] nodataval=ARRAY[0], double precision upperleftx=NULL, double precision upperlefty=NULL, double precision skewx=0, double precision skewy=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, integer width, integer height, double precision gridx, double precision gridy, text pixeltype, double precision value=1, double precision nodataval=0, double precision skewx=0, double precision skewy=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, integer width, integer height, double precision gridx=NULL, double precision gridy=NULL, text[] pixeltype=ARRAY['8BUI'], double precision[] value=ARRAY[1], double precision[] nodataval=ARRAY[0], double precision skewx=0, double precision skewy=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, integer width, integer height, text pixeltype, double precision value=1, double precision nodataval=0, double precision upperleftx=NULL, double precision upperlefty=NULL, double precision skewx=0, double precision skewy=0, boolean touched=false)
;
raster ST_AsRaster(
geometry geom, integer width, integer height, text[] pixeltype, double precision[] value=ARRAY[1], double precision[] nodataval=ARRAY[0], double precision upperleftx=NULL, double precision upperlefty=NULL, double precision skewx=0, double precision skewy=0, boolean touched=false)
;
Converts a PostGIS geometry to a PostGIS raster. The many variants offers three groups of possibilities for setting the alignment and pixelsize of the resulting raster.
The first group, composed of the two first variants, produce a raster having the same alignment (scalex
, scaley
, gridx
and gridy
), pixel type and nodata value as the provided reference raster. You generally pass this reference raster by joining the table containing the geometry with the table containing the reference raster.
The second group, composed of four variants, let you set the dimensions of the raster by providing the parameters of a pixel size (scalex
& scaley
and skewx
& skewy
). The width
& height
of the resulting raster will be adjusted to fit the extent of the geometry. In most cases, you must cast integer scalex
& scaley
arguments to double precision so that PostgreSQL choose the right variant.
The third group, composed of four variants, let you fix the dimensions of the raster by providing the dimensions of the raster (width
& height
). The parameters of the pixel size (scalex
& scaley
and skewx
& skewy
) of the resulting raster will be adjusted to fit the extent of the geometry.
The two first variants of each of those two last groups let you specify the alignment with an arbitrary corner of the alignment grid (gridx
& gridy
) and the two last variants takes the upper left corner (upperleftx
& upperlefty
).
Each group of variant allows producing a one band raster or a multiple bands raster. To produce a multiple bands raster, you must provide an array of pixel types (pixeltype[]
), an array of initial values (value
) and an array of nodata values (nodataval
). If not provided pixeltyped defaults to 8BUI, values to 1 and nodataval to 0.
The output raster will be in the same spatial reference as the source geometry. The only exception is for variants with a reference raster. In this case the resulting raster will get the same SRID as the reference raster.
The optional touched
parameter defaults to false and maps to the GDAL ALL_TOUCHED rasterization option, which determines if pixels touched by lines or polygons will be burned. Not just those on the line render path, or whose center point is within the polygon.
This is particularly useful for rendering jpegs and pngs of geometries directly from the database when using in combination with ST_AsPNG and other ST_AsGDALRaster family of functions.
Availability: 2.0.0 - requires GDAL >= 1.6.0.
Not yet capable of rendering complex geometry types such as curves, TINS, and PolyhedralSurfaces, but should be able too once GDAL can. |
-- this will output a black circle taking up 150 x 150 pixels -- SELECT ST_AsPNG(ST_AsRaster(ST_Buffer(ST_Point(1,5),10),150, 150, '2BUI'));
-- the bands map to RGB bands - the value (118,154,118) - teal -- SELECT ST_AsPNG( ST_AsRaster( ST_Buffer( ST_GeomFromText('LINESTRING(50 50,150 150,150 50)'), 10,'join=bevel'), 200,200,ARRAY['8BUI', '8BUI', '8BUI'], ARRAY[118,154,118], ARRAY[0,0,0]));
ST_BandPixelType, ST_Buffer, ST_GDALDrivers, ST_AsGDALRaster, ST_AsPNG, ST_AsJPEG, ST_SRID
ST_Band — Returns one or more bands of an existing raster as a new raster. Useful for building new rasters from existing rasters.
raster ST_Band(
raster rast, integer[] nbands = ARRAY[1])
;
raster ST_Band(
raster rast, integer nband)
;
raster ST_Band(
raster rast, text nbands, character delimiter=,)
;
Returns one or more bands of an existing raster as a new raster. Useful for building new rasters from existing rasters or export of only selected bands of a raster or rearranging the order of bands in a raster. If no band is specified, band 1 is assumed. Used as a helper function in various functions such as for deleting a band.
For the |
Disponibilità: 2.0
-- Make 2 new rasters: 1 containing band 1 of dummy, second containing band 2 of dummy and then reclassified as a 2BUI SELECT ST_NumBands(rast1) As numb1, ST_BandPixelType(rast1) As pix1, ST_NumBands(rast2) As numb2, ST_BandPixelType(rast2) As pix2 FROM ( SELECT ST_Band(rast) As rast1, ST_Reclass(ST_Band(rast,3), '100-200):1, [200-254:2', '2BUI') As rast2 FROM dummy_rast WHERE rid = 2) As foo; numb1 | pix1 | numb2 | pix2 -------+------+-------+------ 1 | 8BUI | 1 | 2BUI
-- Return bands 2 and 3. Using array cast syntax SELECT ST_NumBands(ST_Band(rast, '{2,3}'::int[])) As num_bands FROM dummy_rast WHERE rid=2; num_bands ---------- 2 -- Return bands 2 and 3. Use array to define bands SELECT ST_NumBands(ST_Band(rast, ARRAY[2,3])) As num_bands FROM dummy_rast WHERE rid=2;
--Make a new raster with 2nd band of original and 1st band repeated twice, and another with just the third band SELECT rast, ST_Band(rast, ARRAY[2,1,1]) As dupe_band, ST_Band(rast, 3) As sing_band FROM samples.than_chunked WHERE rid=35;
ST_MakeEmptyRaster — Returns an empty raster (having no bands) of given dimensions (width & height), upperleft X and Y, pixel size and rotation (scalex, scaley, skewx & skewy) and reference system (srid). If a raster is passed in, returns a new raster with the same size, alignment and SRID. If srid is left out, the spatial ref is set to unknown (0).
raster ST_MakeEmptyRaster(
raster rast)
;
raster ST_MakeEmptyRaster(
integer width, integer height, float8 upperleftx, float8 upperlefty, float8 scalex, float8 scaley, float8 skewx, float8 skewy, integer srid=unknown)
;
raster ST_MakeEmptyRaster(
integer width, integer height, float8 upperleftx, float8 upperlefty, float8 pixelsize)
;
Returns an empty raster (having no band) of given dimensions (width & height) and georeferenced in spatial (or world) coordinates with upper left X (upperleftx), upper left Y (upperlefty), pixel size and rotation (scalex, scaley, skewx & skewy) and reference system (srid).
The last version use a single parameter to specify the pixel size (pixelsize). scalex is set to this argument and scaley is set to the negative value of this argument. skewx and skewy are set to 0.
If an existing raster is passed in, it returns a new raster with the same meta data settings (without the bands).
If no srid is specified it defaults to 0. After you create an empty raster you probably want to add bands to it and maybe edit it. Refer to ST_AddBand to define bands and ST_SetValue to set initial pixel values.
INSERT INTO dummy_rast(rid,rast) VALUES(3, ST_MakeEmptyRaster( 100, 100, 0.0005, 0.0005, 1, 1, 0, 0, 4326) ); --use an existing raster as template for new raster INSERT INTO dummy_rast(rid,rast) SELECT 4, ST_MakeEmptyRaster(rast) FROM dummy_rast WHERE rid = 3; -- output meta data of rasters we just added SELECT rid, (md).* FROM (SELECT rid, ST_MetaData(rast) As md FROM dummy_rast WHERE rid IN(3,4)) As foo; -- output -- rid | upperleftx | upperlefty | width | height | scalex | scaley | skewx | skewy | srid | numbands -----+------------+------------+-------+--------+------------+------------+-------+-------+------+---------- 3 | 0.0005 | 0.0005 | 100 | 100 | 1 | 1 | 0 | 0 | 4326 | 0 4 | 0.0005 | 0.0005 | 100 | 100 | 1 | 1 | 0 | 0 | 4326 | 0
ST_Tile — Returns a set of rasters resulting from the split of the input raster based upon the desired dimensions of the output rasters.
setof raster ST_Tile(
raster rast, int[] nband, integer width, integer height, boolean padwithnodata=FALSE, double precision nodataval=NULL)
;
setof raster ST_Tile(
raster rast, integer nband, integer width, integer height, boolean padwithnodata=FALSE, double precision nodataval=NULL)
;
setof raster ST_Tile(
raster rast, integer width, integer height, boolean padwithnodata=FALSE, double precision nodataval=NULL)
;
Returns a set of rasters resulting from the split of the input raster based upon the desired dimensions of the output rasters.
If padwithnodata
= FALSE, edge tiles on the right and bottom sides of the raster may have different dimensions than the rest of the tiles. If padwithnodata
= TRUE, all tiles will have the same dimensions with the possibility that edge tiles being padded with NODATA values. If raster band(s) do not have NODATA value(s) specified, one can be specified by setting nodataval
.
If a specified band of the input raster is out-of-db, the corresponding band in the output rasters will also be out-of-db. |
Availability: 2.1.0
WITH foo AS ( SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0), 2, '8BUI', 10, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 3, 0, 1, -1, 0, 0, 0), 1, '8BUI', 2, 0), 2, '8BUI', 20, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 6, 0, 1, -1, 0, 0, 0), 1, '8BUI', 3, 0), 2, '8BUI', 30, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, -3, 1, -1, 0, 0, 0), 1, '8BUI', 4, 0), 2, '8BUI', 40, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 3, -3, 1, -1, 0, 0, 0), 1, '8BUI', 5, 0), 2, '8BUI', 50, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 6, -3, 1, -1, 0, 0, 0), 1, '8BUI', 6, 0), 2, '8BUI', 60, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, -6, 1, -1, 0, 0, 0), 1, '8BUI', 7, 0), 2, '8BUI', 70, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 3, -6, 1, -1, 0, 0, 0), 1, '8BUI', 8, 0), 2, '8BUI', 80, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 6, -6, 1, -1, 0, 0, 0), 1, '8BUI', 9, 0), 2, '8BUI', 90, 0) AS rast ), bar AS ( SELECT ST_Union(rast) AS rast FROM foo ), baz AS ( SELECT ST_Tile(rast, 3, 3, TRUE) AS rast FROM bar ) SELECT ST_DumpValues(rast) FROM baz; st_dumpvalues ------------------------------------------ (1,"{{1,1,1},{1,1,1},{1,1,1}}") (2,"{{10,10,10},{10,10,10},{10,10,10}}") (1,"{{2,2,2},{2,2,2},{2,2,2}}") (2,"{{20,20,20},{20,20,20},{20,20,20}}") (1,"{{3,3,3},{3,3,3},{3,3,3}}") (2,"{{30,30,30},{30,30,30},{30,30,30}}") (1,"{{4,4,4},{4,4,4},{4,4,4}}") (2,"{{40,40,40},{40,40,40},{40,40,40}}") (1,"{{5,5,5},{5,5,5},{5,5,5}}") (2,"{{50,50,50},{50,50,50},{50,50,50}}") (1,"{{6,6,6},{6,6,6},{6,6,6}}") (2,"{{60,60,60},{60,60,60},{60,60,60}}") (1,"{{7,7,7},{7,7,7},{7,7,7}}") (2,"{{70,70,70},{70,70,70},{70,70,70}}") (1,"{{8,8,8},{8,8,8},{8,8,8}}") (2,"{{80,80,80},{80,80,80},{80,80,80}}") (1,"{{9,9,9},{9,9,9},{9,9,9}}") (2,"{{90,90,90},{90,90,90},{90,90,90}}") (18 rows)
WITH foo AS ( SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0), 2, '8BUI', 10, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 3, 0, 1, -1, 0, 0, 0), 1, '8BUI', 2, 0), 2, '8BUI', 20, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 6, 0, 1, -1, 0, 0, 0), 1, '8BUI', 3, 0), 2, '8BUI', 30, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, -3, 1, -1, 0, 0, 0), 1, '8BUI', 4, 0), 2, '8BUI', 40, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 3, -3, 1, -1, 0, 0, 0), 1, '8BUI', 5, 0), 2, '8BUI', 50, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 6, -3, 1, -1, 0, 0, 0), 1, '8BUI', 6, 0), 2, '8BUI', 60, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, -6, 1, -1, 0, 0, 0), 1, '8BUI', 7, 0), 2, '8BUI', 70, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 3, -6, 1, -1, 0, 0, 0), 1, '8BUI', 8, 0), 2, '8BUI', 80, 0) AS rast UNION ALL SELECT ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 6, -6, 1, -1, 0, 0, 0), 1, '8BUI', 9, 0), 2, '8BUI', 90, 0) AS rast ), bar AS ( SELECT ST_Union(rast) AS rast FROM foo ), baz AS ( SELECT ST_Tile(rast, 3, 3, 2) AS rast FROM bar ) SELECT ST_DumpValues(rast) FROM baz; st_dumpvalues ------------------------------------------ (1,"{{10,10,10},{10,10,10},{10,10,10}}") (1,"{{20,20,20},{20,20,20},{20,20,20}}") (1,"{{30,30,30},{30,30,30},{30,30,30}}") (1,"{{40,40,40},{40,40,40},{40,40,40}}") (1,"{{50,50,50},{50,50,50},{50,50,50}}") (1,"{{60,60,60},{60,60,60},{60,60,60}}") (1,"{{70,70,70},{70,70,70},{70,70,70}}") (1,"{{80,80,80},{80,80,80},{80,80,80}}") (1,"{{90,90,90},{90,90,90},{90,90,90}}") (9 rows)
ST_Retile — Return a set of configured tiles from an arbitrarily tiled raster coverage.
SETOF raster ST_Retile(
regclass tab, name col, geometry ext, float8 sfx, float8 sfy, int tw, int th, text algo='NearestNeighbor')
;
Return a set of tiles having the specified scale (sfx
, sfy
) and max size (tw
, th
) and covering the specified extent (ext
) with data coming from the specified raster coverage (tab
, col
).
Algorithm options are: 'NearestNeighbor', 'Bilinear', 'Cubic', 'CubicSpline', and 'Lanczos'. Refer to: GDAL Warp resampling methods for more details.
Availability: 2.2.0
ST_FromGDALRaster — Returns a raster from a supported GDAL raster file.
raster ST_FromGDALRaster(
bytea gdaldata, integer srid=NULL)
;
Returns a raster from a supported GDAL raster file. gdaldata
is of type bytea and should be the contents of the GDAL raster file.
If srid
is NULL, the function will try to automatically assign the SRID from the GDAL raster. If srid
is provided, the value provided will override any automatically assigned SRID.
Availability: 2.1.0
WITH foo AS ( SELECT ST_AsPNG(ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 0.1, -0.1, 0, 0, 4326), 1, '8BUI', 1, 0), 2, '8BUI', 2, 0), 3, '8BUI', 3, 0)) AS png ), bar AS ( SELECT 1 AS rid, ST_FromGDALRaster(png) AS rast FROM foo UNION ALL SELECT 2 AS rid, ST_FromGDALRaster(png, 3310) AS rast FROM foo ) SELECT rid, ST_Metadata(rast) AS metadata, ST_SummaryStats(rast, 1) AS stats1, ST_SummaryStats(rast, 2) AS stats2, ST_SummaryStats(rast, 3) AS stats3 FROM bar ORDER BY rid; rid | metadata | stats1 | stats2 | stats3 -----+---------------------------+---------------+---------------+---------------- 1 | (0,0,2,2,1,-1,0,0,0,3) | (4,4,1,0,1,1) | (4,8,2,0,2,2) | (4,12,3,0,3,3) 2 | (0,0,2,2,1,-1,0,0,3310,3) | (4,4,1,0,1,1) | (4,8,2,0,2,2) | (4,12,3,0,3,3) (2 rows)
ST_GeoReference — Returns the georeference meta data in GDAL or ESRI format as commonly seen in a world file. Default is GDAL.
text ST_GeoReference(
raster rast, text format=GDAL)
;
Returns the georeference meta data including carriage return in GDAL or ESRI format as commonly seen in a world file. Default is GDAL if no type specified. type is string 'GDAL' or 'ESRI'.
Difference between format representations is as follows:
GDAL
:
scalex skewy skewx scaley upperleftx upperlefty
ESRI
:
scalex skewy skewx scaley upperleftx + scalex*0.5 upperlefty + scaley*0.5
SELECT ST_GeoReference(rast, 'ESRI') As esri_ref, ST_GeoReference(rast, 'GDAL') As gdal_ref FROM dummy_rast WHERE rid=1; esri_ref | gdal_ref --------------+-------------- 2.0000000000 | 2.0000000000 0.0000000000 : 0.0000000000 0.0000000000 : 0.0000000000 3.0000000000 : 3.0000000000 1.5000000000 : 0.5000000000 2.0000000000 : 0.5000000000
ST_Height — Returns the height of the raster in pixels.
integer ST_Height(
raster rast)
;
ST_IsEmpty — Returns true if the raster is empty (width = 0 and height = 0). Otherwise, returns false.
boolean ST_IsEmpty(
raster rast)
;
Returns true if the raster is empty (width = 0 and height = 0). Otherwise, returns false.
Disponibilità: 2.0
ST_MemSize — Returns the amount of space (in bytes) the raster takes.
integer ST_MemSize(
raster rast)
;
Returns the amount of space (in bytes) the raster takes.
This is a nice compliment to PostgreSQL built in functions pg_column_size, pg_size_pretty, pg_relation_size, pg_total_relation_size.
pg_relation_size which gives the byte size of a table may return byte size lower than ST_MemSize. This is because pg_relation_size does not add toasted table contribution and large geometries are stored in TOAST tables. pg_column_size might return lower because it returns the compressed size. pg_total_relation_size - includes, the table, the toasted tables, and the indexes. |
Availability: 2.2.0
ST_MetaData — Returns basic meta data about a raster object such as pixel size, rotation (skew), upper, lower left, etc.
record ST_MetaData(
raster rast)
;
Returns basic meta data about a raster object such as pixel size, rotation (skew), upper, lower left, etc. Columns returned: upperleftx | upperlefty | width | height | scalex | scaley | skewx | skewy | srid | numbands
SELECT rid, (foo.md).* FROM (SELECT rid, ST_MetaData(rast) As md FROM dummy_rast) As foo; rid | upperleftx | upperlefty | width | height | scalex | scaley | skewx | skewy | srid | numbands ----+------------+------------+-------+--------+--------+-----------+-------+-------+------+------- 1 | 0.5 | 0.5 | 10 | 20 | 2 | 3 | 0 | 0 | 0 | 0 2 | 3427927.75 | 5793244 | 5 | 5 | 0.05 | -0.05 | 0 | 0 | 0 | 3
ST_NumBands — Returns the number of bands in the raster object.
integer ST_NumBands(
raster rast)
;
ST_PixelHeight — Returns the pixel height in geometric units of the spatial reference system.
double precision ST_PixelHeight(
raster rast)
;
Returns the height of a pixel in geometric units of the spatial reference system. In the common case where there is no skew, the pixel height is just the scale ratio between geometric coordinates and raster pixels.
Refer to ST_PixelWidth for a diagrammatic visualization of the relationship.
SELECT ST_Height(rast) As rastheight, ST_PixelHeight(rast) As pixheight, ST_ScaleX(rast) As scalex, ST_ScaleY(rast) As scaley, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy FROM dummy_rast; rastheight | pixheight | scalex | scaley | skewx | skewy ------------+-----------+--------+--------+-------+---------- 20 | 3 | 2 | 3 | 0 | 0 5 | 0.05 | 0.05 | -0.05 | 0 | 0
SELECT ST_Height(rast) As rastheight, ST_PixelHeight(rast) As pixheight, ST_ScaleX(rast) As scalex, ST_ScaleY(rast) As scaley, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy FROM (SELECT ST_SetSKew(rast,0.5,0.5) As rast FROM dummy_rast) As skewed; rastheight | pixheight | scalex | scaley | skewx | skewy -----------+-------------------+--------+--------+-------+---------- 20 | 3.04138126514911 | 2 | 3 | 0.5 | 0.5 5 | 0.502493781056044 | 0.05 | -0.05 | 0.5 | 0.5
ST_PixelWidth — Returns the pixel width in geometric units of the spatial reference system.
double precision ST_PixelWidth(
raster rast)
;
Returns the width of a pixel in geometric units of the spatial reference system. In the common case where there is no skew, the pixel width is just the scale ratio between geometric coordinates and raster pixels.
The following diagram demonstrates the relationship:
SELECT ST_Width(rast) As rastwidth, ST_PixelWidth(rast) As pixwidth, ST_ScaleX(rast) As scalex, ST_ScaleY(rast) As scaley, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy FROM dummy_rast; rastwidth | pixwidth | scalex | scaley | skewx | skewy -----------+----------+--------+--------+-------+---------- 10 | 2 | 2 | 3 | 0 | 0 5 | 0.05 | 0.05 | -0.05 | 0 | 0
SELECT ST_Width(rast) As rastwidth, ST_PixelWidth(rast) As pixwidth, ST_ScaleX(rast) As scalex, ST_ScaleY(rast) As scaley, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy FROM (SELECT ST_SetSkew(rast,0.5,0.5) As rast FROM dummy_rast) As skewed; rastwidth | pixwidth | scalex | scaley | skewx | skewy -----------+-------------------+--------+--------+-------+---------- 10 | 2.06155281280883 | 2 | 3 | 0.5 | 0.5 5 | 0.502493781056044 | 0.05 | -0.05 | 0.5 | 0.5
ST_ScaleX — Returns the X component of the pixel width in units of coordinate reference system.
float8 ST_ScaleX(
raster rast)
;
Returns the X component of the pixel width in units of coordinate reference system. Refer to World File for more details.
Changed: 2.0.0. In WKTRaster versions this was called ST_PixelSizeX.
ST_ScaleY — Returns the Y component of the pixel height in units of coordinate reference system.
float8 ST_ScaleY(
raster rast)
;
Returns the Y component of the pixel height in units of coordinate reference system. May be negative. Refer to World File for more details.
Changed: 2.0.0. In WKTRaster versions this was called ST_PixelSizeY.
ST_RasterToWorldCoord — Returns the raster's upper left corner as geometric X and Y (longitude and latitude) given a column and row. Column and row starts at 1.
record ST_RasterToWorldCoord(
raster rast, integer xcolumn, integer yrow)
;
Returns the upper left corner as geometric X and Y (longitude and latitude) given a column and row. Returned X and Y are in geometric units of the georeferenced raster. Numbering of column and row starts at 1 but if either parameter is passed a zero, a negative number or a number greater than the respective dimension of the raster, it will return coordinates outside of the raster assuming the raster's grid is applicable outside the raster's bounds.
Availability: 2.1.0
-- non-skewed raster SELECT rid, (ST_RasterToWorldCoord(rast,1, 1)).*, (ST_RasterToWorldCoord(rast,2, 2)).* FROM dummy_rast rid | longitude | latitude | longitude | latitude -----+------------+----------+-----------+------------ 1 | 0.5 | 0.5 | 2.5 | 3.5 2 | 3427927.75 | 5793244 | 3427927.8 | 5793243.95
-- skewed raster SELECT rid, (ST_RasterToWorldCoord(rast, 1, 1)).*, (ST_RasterToWorldCoord(rast, 2, 3)).* FROM ( SELECT rid, ST_SetSkew(rast, 100.5, 0) As rast FROM dummy_rast ) As foo rid | longitude | latitude | longitude | latitude -----+------------+----------+-----------+----------- 1 | 0.5 | 0.5 | 203.5 | 6.5 2 | 3427927.75 | 5793244 | 3428128.8 | 5793243.9
ST_RasterToWorldCoordX — Returns the geometric X coordinate upper left of a raster, column and row. Numbering of columns and rows starts at 1.
float8 ST_RasterToWorldCoordX(
raster rast, integer xcolumn)
;
float8 ST_RasterToWorldCoordX(
raster rast, integer xcolumn, integer yrow)
;
Returns the upper left X coordinate of a raster column row in geometric units of the georeferenced raster. Numbering of columns and rows starts at 1 but if you pass in a negative number or number higher than number of columns in raster, it will give you coordinates outside of the raster file to left or right with the assumption that the skew and pixel sizes are same as selected raster.
For non-skewed rasters, providing the X column is sufficient. For skewed rasters, the georeferenced coordinate is a function of the ST_ScaleX and ST_SkewX and row and column. An error will be raised if you give just the X column for a skewed raster. |
Changed: 2.1.0 In prior versions, this was called ST_Raster2WorldCoordX
-- non-skewed raster providing column is sufficient SELECT rid, ST_RasterToWorldCoordX(rast,1) As x1coord, ST_RasterToWorldCoordX(rast,2) As x2coord, ST_ScaleX(rast) As pixelx FROM dummy_rast; rid | x1coord | x2coord | pixelx -----+------------+-----------+-------- 1 | 0.5 | 2.5 | 2 2 | 3427927.75 | 3427927.8 | 0.05
-- for fun lets skew it SELECT rid, ST_RasterToWorldCoordX(rast, 1, 1) As x1coord, ST_RasterToWorldCoordX(rast, 2, 3) As x2coord, ST_ScaleX(rast) As pixelx FROM (SELECT rid, ST_SetSkew(rast, 100.5, 0) As rast FROM dummy_rast) As foo; rid | x1coord | x2coord | pixelx -----+------------+-----------+-------- 1 | 0.5 | 203.5 | 2 2 | 3427927.75 | 3428128.8 | 0.05
ST_RasterToWorldCoordY — Returns the geometric Y coordinate upper left corner of a raster, column and row. Numbering of columns and rows starts at 1.
float8 ST_RasterToWorldCoordY(
raster rast, integer yrow)
;
float8 ST_RasterToWorldCoordY(
raster rast, integer xcolumn, integer yrow)
;
Returns the upper left Y coordinate of a raster column row in geometric units of the georeferenced raster. Numbering of columns and rows starts at 1 but if you pass in a negative number or number higher than number of columns/rows in raster, it will give you coordinates outside of the raster file to left or right with the assumption that the skew and pixel sizes are same as selected raster tile.
For non-skewed rasters, providing the Y column is sufficient. For skewed rasters, the georeferenced coordinate is a function of the ST_ScaleY and ST_SkewY and row and column. An error will be raised if you give just the Y row for a skewed raster. |
Changed: 2.1.0 In prior versions, this was called ST_Raster2WorldCoordY
-- non-skewed raster providing row is sufficient SELECT rid, ST_RasterToWorldCoordY(rast,1) As y1coord, ST_RasterToWorldCoordY(rast,3) As y2coord, ST_ScaleY(rast) As pixely FROM dummy_rast; rid | y1coord | y2coord | pixely -----+---------+-----------+-------- 1 | 0.5 | 6.5 | 3 2 | 5793244 | 5793243.9 | -0.05
-- for fun lets skew it SELECT rid, ST_RasterToWorldCoordY(rast,1,1) As y1coord, ST_RasterToWorldCoordY(rast,2,3) As y2coord, ST_ScaleY(rast) As pixely FROM (SELECT rid, ST_SetSkew(rast,0,100.5) As rast FROM dummy_rast) As foo; rid | y1coord | y2coord | pixely -----+---------+-----------+-------- 1 | 0.5 | 107 | 3 2 | 5793244 | 5793344.4 | -0.05
ST_Rotation — Returns the rotation of the raster in radian.
float8 ST_Rotation(
raster rast)
;
Returns the uniform rotation of the raster in radian. If a raster does not have uniform rotation, NaN is returned. Refer to World File for more details.
ST_SkewX — Returns the georeference X skew (or rotation parameter).
float8 ST_SkewX(
raster rast)
;
Returns the georeference X skew (or rotation parameter). Refer to World File for more details.
SELECT rid, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy, ST_GeoReference(rast) as georef FROM dummy_rast; rid | skewx | skewy | georef -----+-------+-------+-------------------- 1 | 0 | 0 | 2.0000000000 : 0.0000000000 : 0.0000000000 : 3.0000000000 : 0.5000000000 : 0.5000000000 : 2 | 0 | 0 | 0.0500000000 : 0.0000000000 : 0.0000000000 : -0.0500000000 : 3427927.7500000000 : 5793244.0000000000
ST_SkewY — Returns the georeference Y skew (or rotation parameter).
float8 ST_SkewY(
raster rast)
;
Returns the georeference Y skew (or rotation parameter). Refer to World File for more details.
SELECT rid, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy, ST_GeoReference(rast) as georef FROM dummy_rast; rid | skewx | skewy | georef -----+-------+-------+-------------------- 1 | 0 | 0 | 2.0000000000 : 0.0000000000 : 0.0000000000 : 3.0000000000 : 0.5000000000 : 0.5000000000 : 2 | 0 | 0 | 0.0500000000 : 0.0000000000 : 0.0000000000 : -0.0500000000 : 3427927.7500000000 : 5793244.0000000000
ST_SRID — Returns the spatial reference identifier of the raster as defined in spatial_ref_sys table.
integer ST_SRID(
raster rast)
;
ST_Summary — Returns a text summary of the contents of the raster.
text ST_Summary(
raster rast)
;
SELECT ST_Summary( ST_AddBand( ST_AddBand( ST_AddBand( ST_MakeEmptyRaster(10, 10, 0, 0, 1, -1, 0, 0, 0) , 1, '8BUI', 1, 0 ) , 2, '32BF', 0, -9999 ) , 3, '16BSI', 0, NULL ) ); st_summary ------------------------------------------------------------------ Raster of 10x10 pixels has 3 bands and extent of BOX(0 -10,10 0)+ band 1 of pixtype 8BUI is in-db with NODATA value of 0 + band 2 of pixtype 32BF is in-db with NODATA value of -9999 + band 3 of pixtype 16BSI is in-db with no NODATA value (1 row)
ST_UpperLeftX — Returns the upper left X coordinate of raster in projected spatial ref.
float8 ST_UpperLeftX(
raster rast)
;
ST_UpperLeftY — Returns the upper left Y coordinate of raster in projected spatial ref.
float8 ST_UpperLeftY(
raster rast)
;
ST_WorldToRasterCoord — Returns the upper left corner as column and row given geometric X and Y (longitude and latitude) or a point geometry expressed in the spatial reference coordinate system of the raster.
record ST_WorldToRasterCoord(
raster rast, geometry pt)
;
record ST_WorldToRasterCoord(
raster rast, double precision longitude, double precision latitude)
;
Returns the upper left corner as column and row given geometric X and Y (longitude and latitude) or a point geometry. This function works regardless of whether or not the geometric X and Y or point geometry is outside the extent of the raster. Geometric X and Y must be expressed in the spatial reference coordinate system of the raster.
Availability: 2.1.0
SELECT rid, (ST_WorldToRasterCoord(rast,3427927.8,20.5)).*, (ST_WorldToRasterCoord(rast,ST_GeomFromText('POINT(3427927.8 20.5)',ST_SRID(rast)))).* FROM dummy_rast; rid | columnx | rowy | columnx | rowy -----+---------+-----------+---------+----------- 1 | 1713964 | 7 | 1713964 | 7 2 | 2 | 115864471 | 2 | 115864471
ST_WorldToRasterCoordX — Returns the column in the raster of the point geometry (pt) or a X and Y world coordinate (xw, yw) represented in world spatial reference system of raster.
integer ST_WorldToRasterCoordX(
raster rast, geometry pt)
;
integer ST_WorldToRasterCoordX(
raster rast, double precision xw)
;
integer ST_WorldToRasterCoordX(
raster rast, double precision xw, double precision yw)
;
Returns the column in the raster of the point geometry (pt) or a X and Y world coordinate (xw, yw). A point, or (both xw and yw world coordinates are required if a raster is skewed). If a raster is not skewed then xw is sufficient. World coordinates are in the spatial reference coordinate system of the raster.
Changed: 2.1.0 In prior versions, this was called ST_World2RasterCoordX
SELECT rid, ST_WorldToRasterCoordX(rast,3427927.8) As xcoord, ST_WorldToRasterCoordX(rast,3427927.8,20.5) As xcoord_xwyw, ST_WorldToRasterCoordX(rast,ST_GeomFromText('POINT(3427927.8 20.5)',ST_SRID(rast))) As ptxcoord FROM dummy_rast; rid | xcoord | xcoord_xwyw | ptxcoord -----+---------+---------+---------- 1 | 1713964 | 1713964 | 1713964 2 | 1 | 1 | 1
ST_WorldToRasterCoordY — Returns the row in the raster of the point geometry (pt) or a X and Y world coordinate (xw, yw) represented in world spatial reference system of raster.
integer ST_WorldToRasterCoordY(
raster rast, geometry pt)
;
integer ST_WorldToRasterCoordY(
raster rast, double precision xw)
;
integer ST_WorldToRasterCoordY(
raster rast, double precision xw, double precision yw)
;
Returns the row in the raster of the point geometry (pt) or a X and Y world coordinate (xw, yw). A point, or (both xw and yw world coordinates are required if a raster is skewed). If a raster is not skewed then xw is sufficient. World coordinates are in the spatial reference coordinate system of the raster.
Changed: 2.1.0 In prior versions, this was called ST_World2RasterCoordY
SELECT rid, ST_WorldToRasterCoordY(rast,20.5) As ycoord, ST_WorldToRasterCoordY(rast,3427927.8,20.5) As ycoord_xwyw, ST_WorldToRasterCoordY(rast,ST_GeomFromText('POINT(3427927.8 20.5)',ST_SRID(rast))) As ptycoord FROM dummy_rast; rid | ycoord | ycoord_xwyw | ptycoord -----+-----------+-------------+----------- 1 | 7 | 7 | 7 2 | 115864471 | 115864471 | 115864471
ST_BandMetaData — Returns basic meta data for a specific raster band. band num 1 is assumed if none-specified.
record ST_BandMetaData(
raster rast, integer bandnum=1)
;
Returns basic meta data about a raster band. Columns returned pixeltype | nodatavalue | isoutdb | path.
If raster contains no bands then an error is thrown. |
If band has no NODATA value, nodatavalue will be NULL. |
ST_BandNoDataValue — Returns the value in a given band that represents no data. If no band num 1 is assumed.
double precision ST_BandNoDataValue(
raster rast, integer bandnum=1)
;
ST_BandIsNoData — Returns true if the band is filled with only nodata values.
boolean ST_BandIsNoData(
raster rast, integer band, boolean forceChecking=true)
;
boolean ST_BandIsNoData(
raster rast, boolean forceChecking=true)
;
Returns true if the band is filled with only nodata values. Band 1 is assumed if not specified. If the last argument is TRUE, the entire band is checked pixel by pixel. Otherwise, the function simply returns the value of the isnodata flag for the band. The default value for this parameter is FALSE, if not specified.
Disponibilità: 2.0
If the flag is dirty (this is, the result is different using TRUE as last parameter and not using it) you should update the raster to set this flag to true, by using ST_SetBandIsNodata(), or ST_SetBandNodataValue() with TRUE as last argument. See ST_SetBandIsNoData. |
-- Create dummy table with one raster column create table dummy_rast (rid integer, rast raster); -- Add raster with two bands, one pixel/band. In the first band, nodatavalue = pixel value = 3. -- In the second band, nodatavalue = 13, pixel value = 4 insert into dummy_rast values(1, ( '01' -- little endian (uint8 ndr) || '0000' -- version (uint16 0) || '0200' -- nBands (uint16 0) || '17263529ED684A3F' -- scaleX (float64 0.000805965234044584) || 'F9253529ED684ABF' -- scaleY (float64 -0.00080596523404458) || '1C9F33CE69E352C0' -- ipX (float64 -75.5533328537098) || '718F0E9A27A44840' -- ipY (float64 49.2824585505576) || 'ED50EB853EC32B3F' -- skewX (float64 0.000211812383858707) || '7550EB853EC32B3F' -- skewY (float64 0.000211812383858704) || 'E6100000' -- SRID (int32 4326) || '0100' -- width (uint16 1) || '0100' -- height (uint16 1) || '6' -- hasnodatavalue and isnodata value set to true. || '2' -- first band type (4BUI) || '03' -- novalue==3 || '03' -- pixel(0,0)==3 (same that nodata) || '0' -- hasnodatavalue set to false || '5' -- second band type (16BSI) || '0D00' -- novalue==13 || '0400' -- pixel(0,0)==4 )::raster ); select st_bandisnodata(rast, 1) from dummy_rast where rid = 1; -- Expected true select st_bandisnodata(rast, 2) from dummy_rast where rid = 1; -- Expected false
ST_BandPath — Returns system file path to a band stored in file system. If no bandnum specified, 1 is assumed.
text ST_BandPath(
raster rast, integer bandnum=1)
;
ST_BandPixelType — Returns the type of pixel for given band. If no bandnum specified, 1 is assumed.
text ST_BandPixelType(
raster rast, integer bandnum=1)
;
Returns the value that represents no data for the band
There are 11 pixel types. Pixel Types supported are as follows:
1BB - 1-bit boolean
2BUI - 2-bit unsigned integer
4BUI - 4-bit unsigned integer
8BSI - 8-bit signed integer
8BUI - 8-bit unsigned integer
16BSI - 16-bit signed integer
16BUI - 16-bit unsigned integer
32BSI - 32-bit signed integer
32BUI - 32-bit unsigned integer
32BF - 32-bit float
64BF - 64-bit float
ST_HasNoBand — Returns true if there is no band with given band number. If no band number is specified, then band number 1 is assumed.
boolean ST_HasNoBand(
raster rast, integer bandnum=1)
;
Returns true if there is no band with given band number. If no band number is specified, then band number 1 is assumed.
Disponibilità: 2.0
exclude_nodata_value
is set to false, then all pixels include nodata
pixels are considered to intersect and return value. If exclude_nodata_value
is not passed in then reads it from metadata of raster.NODATA
value of a given band's pixel specified by a columnx and rowy or a geometric point expressed in the same spatial reference coordinate system as the raster. NODATA
values around a given band's pixel specified by either a columnX and rowY or a geometric point expressed in the same spatial reference coordinate system as the raster. ST_PixelAsPolygon — Returns the polygon geometry that bounds the pixel for a particular row and column.
geometry ST_PixelAsPolygon(
raster rast, integer columnx, integer rowy)
;
Returns the polygon geometry that bounds the pixel for a particular row and column.
Disponibilità: 2.0
-- get raster pixel polygon SELECT i,j, ST_AsText(ST_PixelAsPolygon(foo.rast, i,j)) As b1pgeom FROM dummy_rast As foo CROSS JOIN generate_series(1,2) As i CROSS JOIN generate_series(1,1) As j WHERE rid=2; i | j | b1pgeom ---+---+----------------------------------------------------------------------------- 1 | 1 | POLYGON((3427927.75 5793244,3427927.8 5793244,3427927.8 5793243.95,... 2 | 1 | POLYGON((3427927.8 5793244,3427927.85 5793244,3427927.85 5793243.95, ..
ST_PixelAsPolygons — Returns the polygon geometry that bounds every pixel of a raster band along with the value, the X and the Y raster coordinates of each pixel.
setof record ST_PixelAsPolygons(
raster rast, integer band=1, boolean exclude_nodata_value=TRUE)
;
Returns the polygon geometry that bounds every pixel of a raster band along with the value (double precision), the X and the Y raster coordinates (integers) of each pixel.
ST_PixelAsPolygons returns one polygon geometry for every pixel. This is different than ST_DumpAsPolygons where each geometry represents one or more pixels with the same pixel value. |
When exclude_nodata_value = TRUE, only those pixels whose values are not NODATA are returned as polygons. |
Disponibilità: 2.0
Enhanced: 2.1.0 exclude_nodata_value optional argument was added.
Changed: 2.1.1 Changed behavior of exclude_nodata_value.
-- get raster pixel polygon SELECT (gv).x, (gv).y, (gv).val, ST_AsText((gv).geom) geom FROM (SELECT ST_PixelAsPolygons( ST_SetValue(ST_SetValue(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 0.001, -0.001, 0.001, 0.001, 4269), '8BUI'::text, 1, 0), 2, 2, 10), 1, 1, NULL) ) gv ) foo; x | y | val | geom ---+---+----------------------------------------------------------------------------- 1 | 1 | | POLYGON((0 0,0.001 0.001,0.002 0,0.001 -0.001,0 0)) 1 | 2 | 1 | POLYGON((0.001 -0.001,0.002 0,0.003 -0.001,0.002 -0.002,0.001 -0.001)) 2 | 1 | 1 | POLYGON((0.001 0.001,0.002 0.002,0.003 0.001,0.002 0,0.001 0.001)) 2 | 2 | 10 | POLYGON((0.002 0,0.003 0.001,0.004 0,0.003 -0.001,0.002 0))
ST_PixelAsPoint — Returns a point geometry of the pixel's upper-left corner.
geometry ST_PixelAsPoint(
raster rast, integer columnx, integer rowy)
;
ST_PixelAsPoints — Returns a point geometry for each pixel of a raster band along with the value, the X and the Y raster coordinates of each pixel. The coordinates of the point geometry are of the pixel's upper-left corner.
geometry ST_PixelAsPoints(
raster rast, integer band=1, boolean exclude_nodata_value=TRUE)
;
Returns a point geometry for each pixel of a raster band along with the value, the X and the Y raster coordinates of each pixel. The coordinates of the point geometry are of the pixel's upper-left corner.
When exclude_nodata_value = TRUE, only those pixels whose values are not NODATA are returned as points. |
Availability: 2.1.0
Changed: 2.1.1 Changed behavior of exclude_nodata_value.
SELECT x, y, val, ST_AsText(geom) FROM (SELECT (ST_PixelAsPoints(rast, 1)).* FROM dummy_rast WHERE rid = 2) foo; x | y | val | st_astext ---+---+-----+------------------------------ 1 | 1 | 253 | POINT(3427927.75 5793244) 2 | 1 | 254 | POINT(3427927.8 5793244) 3 | 1 | 253 | POINT(3427927.85 5793244) 4 | 1 | 254 | POINT(3427927.9 5793244) 5 | 1 | 254 | POINT(3427927.95 5793244) 1 | 2 | 253 | POINT(3427927.75 5793243.95) 2 | 2 | 254 | POINT(3427927.8 5793243.95) 3 | 2 | 254 | POINT(3427927.85 5793243.95) 4 | 2 | 253 | POINT(3427927.9 5793243.95) 5 | 2 | 249 | POINT(3427927.95 5793243.95) 1 | 3 | 250 | POINT(3427927.75 5793243.9) 2 | 3 | 254 | POINT(3427927.8 5793243.9) 3 | 3 | 254 | POINT(3427927.85 5793243.9) 4 | 3 | 252 | POINT(3427927.9 5793243.9) 5 | 3 | 249 | POINT(3427927.95 5793243.9) 1 | 4 | 251 | POINT(3427927.75 5793243.85) 2 | 4 | 253 | POINT(3427927.8 5793243.85) 3 | 4 | 254 | POINT(3427927.85 5793243.85) 4 | 4 | 254 | POINT(3427927.9 5793243.85) 5 | 4 | 253 | POINT(3427927.95 5793243.85) 1 | 5 | 252 | POINT(3427927.75 5793243.8) 2 | 5 | 250 | POINT(3427927.8 5793243.8) 3 | 5 | 254 | POINT(3427927.85 5793243.8) 4 | 5 | 254 | POINT(3427927.9 5793243.8) 5 | 5 | 254 | POINT(3427927.95 5793243.8)
ST_PixelAsCentroid — Returns the centroid (point geometry) of the area represented by a pixel.
geometry ST_PixelAsCentroid(
raster rast, integer columnx, integer rowy)
;
Returns the centroid (point geometry) of the area represented by a pixel.
Availability: 2.1.0
ST_PixelAsCentroids — Returns the centroid (point geometry) for each pixel of a raster band along with the value, the X and the Y raster coordinates of each pixel. The point geometry is the centroid of the area represented by a pixel.
geometry ST_PixelAsCentroids(
raster rast, integer band=1, boolean exclude_nodata_value=TRUE)
;
Returns the centroid (point geometry) for each pixel of a raster band along with the value, the X and the Y raster coordinates of each pixel. The point geometry is the centroid of the area represented by a pixel.
When exclude_nodata_value = TRUE, only those pixels whose values are not NODATA are returned as points. |
Availability: 2.1.0
Changed: 2.1.1 Changed behavior of exclude_nodata_value.
SELECT x, y, val, ST_AsText(geom) FROM (SELECT (ST_PixelAsCentroids(rast, 1)).* FROM dummy_rast WHERE rid = 2) foo; x | y | val | st_astext ---+---+-----+-------------------------------- 1 | 1 | 253 | POINT(3427927.775 5793243.975) 2 | 1 | 254 | POINT(3427927.825 5793243.975) 3 | 1 | 253 | POINT(3427927.875 5793243.975) 4 | 1 | 254 | POINT(3427927.925 5793243.975) 5 | 1 | 254 | POINT(3427927.975 5793243.975) 1 | 2 | 253 | POINT(3427927.775 5793243.925) 2 | 2 | 254 | POINT(3427927.825 5793243.925) 3 | 2 | 254 | POINT(3427927.875 5793243.925) 4 | 2 | 253 | POINT(3427927.925 5793243.925) 5 | 2 | 249 | POINT(3427927.975 5793243.925) 1 | 3 | 250 | POINT(3427927.775 5793243.875) 2 | 3 | 254 | POINT(3427927.825 5793243.875) 3 | 3 | 254 | POINT(3427927.875 5793243.875) 4 | 3 | 252 | POINT(3427927.925 5793243.875) 5 | 3 | 249 | POINT(3427927.975 5793243.875) 1 | 4 | 251 | POINT(3427927.775 5793243.825) 2 | 4 | 253 | POINT(3427927.825 5793243.825) 3 | 4 | 254 | POINT(3427927.875 5793243.825) 4 | 4 | 254 | POINT(3427927.925 5793243.825) 5 | 4 | 253 | POINT(3427927.975 5793243.825) 1 | 5 | 252 | POINT(3427927.775 5793243.775) 2 | 5 | 250 | POINT(3427927.825 5793243.775) 3 | 5 | 254 | POINT(3427927.875 5793243.775) 4 | 5 | 254 | POINT(3427927.925 5793243.775) 5 | 5 | 254 | POINT(3427927.975 5793243.775)
ST_Value — Returns the value of a given band in a given columnx, rowy pixel or at a particular geometric point. Band numbers start at 1 and assumed to be 1 if not specified. If exclude_nodata_value
is set to false, then all pixels include nodata
pixels are considered to intersect and return value. If exclude_nodata_value
is not passed in then reads it from metadata of raster.
double precision ST_Value(
raster rast, geometry pt, boolean exclude_nodata_value=true)
;
double precision ST_Value(
raster rast, integer bandnum, geometry pt, boolean exclude_nodata_value=true)
;
double precision ST_Value(
raster rast, integer columnx, integer rowy, boolean exclude_nodata_value=true)
;
double precision ST_Value(
raster rast, integer bandnum, integer columnx, integer rowy, boolean exclude_nodata_value=true)
;
Returns the value of a given band in a given columnx, rowy pixel or at a given geometry point. Band numbers start at 1 and band is assumed to be 1 if not specified. If exclude_nodata_value
is set to true, then only non nodata
pixels are considered. If exclude_nodata_value
is set to false, then all pixels are considered.
Enhanced: 2.0.0 exclude_nodata_value optional argument was added.
-- get raster values at particular postgis geometry points -- the srid of your geometry should be same as for your raster SELECT rid, ST_Value(rast, foo.pt_geom) As b1pval, ST_Value(rast, 2, foo.pt_geom) As b2pval FROM dummy_rast CROSS JOIN (SELECT ST_SetSRID(ST_Point(3427927.77, 5793243.76), 0) As pt_geom) As foo WHERE rid=2; rid | b1pval | b2pval -----+--------+-------- 2 | 252 | 79 -- general fictitious example using a real table SELECT rid, ST_Value(rast, 3, sometable.geom) As b3pval FROM sometable WHERE ST_Intersects(rast,sometable.geom);
SELECT rid, ST_Value(rast, 1, 1, 1) As b1pval, ST_Value(rast, 2, 1, 1) As b2pval, ST_Value(rast, 3, 1, 1) As b3pval FROM dummy_rast WHERE rid=2; rid | b1pval | b2pval | b3pval -----+--------+--------+-------- 2 | 253 | 78 | 70
--- Get all values in bands 1,2,3 of each pixel -- SELECT x, y, ST_Value(rast, 1, x, y) As b1val, ST_Value(rast, 2, x, y) As b2val, ST_Value(rast, 3, x, y) As b3val FROM dummy_rast CROSS JOIN generate_series(1, 1000) As x CROSS JOIN generate_series(1, 1000) As y WHERE rid = 2 AND x <= ST_Width(rast) AND y <= ST_Height(rast); x | y | b1val | b2val | b3val ---+---+-------+-------+------- 1 | 1 | 253 | 78 | 70 1 | 2 | 253 | 96 | 80 1 | 3 | 250 | 99 | 90 1 | 4 | 251 | 89 | 77 1 | 5 | 252 | 79 | 62 2 | 1 | 254 | 98 | 86 2 | 2 | 254 | 118 | 108 : :
--- Get all values in bands 1,2,3 of each pixel same as above but returning the upper left point point of each pixel -- SELECT ST_AsText(ST_SetSRID( ST_Point(ST_UpperLeftX(rast) + ST_ScaleX(rast)*x, ST_UpperLeftY(rast) + ST_ScaleY(rast)*y), ST_SRID(rast))) As uplpt , ST_Value(rast, 1, x, y) As b1val, ST_Value(rast, 2, x, y) As b2val, ST_Value(rast, 3, x, y) As b3val FROM dummy_rast CROSS JOIN generate_series(1,1000) As x CROSS JOIN generate_series(1,1000) As y WHERE rid = 2 AND x <= ST_Width(rast) AND y <= ST_Height(rast); uplpt | b1val | b2val | b3val -----------------------------+-------+-------+------- POINT(3427929.25 5793245.5) | 253 | 78 | 70 POINT(3427929.25 5793247) | 253 | 96 | 80 POINT(3427929.25 5793248.5) | 250 | 99 | 90 :
--- Get a polygon formed by union of all pixels that fall in a particular value range and intersect particular polygon -- SELECT ST_AsText(ST_Union(pixpolyg)) As shadow FROM (SELECT ST_Translate(ST_MakeEnvelope( ST_UpperLeftX(rast), ST_UpperLeftY(rast), ST_UpperLeftX(rast) + ST_ScaleX(rast), ST_UpperLeftY(rast) + ST_ScaleY(rast), 0 ), ST_ScaleX(rast)*x, ST_ScaleY(rast)*y ) As pixpolyg, ST_Value(rast, 2, x, y) As b2val FROM dummy_rast CROSS JOIN generate_series(1,1000) As x CROSS JOIN generate_series(1,1000) As y WHERE rid = 2 AND x <= ST_Width(rast) AND y <= ST_Height(rast)) As foo WHERE ST_Intersects( pixpolyg, ST_GeomFromText('POLYGON((3427928 5793244,3427927.75 5793243.75,3427928 5793243.75,3427928 5793244))',0) ) AND b2val != 254; shadow ------------------------------------------------------------------------------------ MULTIPOLYGON(((3427928 5793243.9,3427928 5793243.85,3427927.95 5793243.85,3427927.95 5793243.9, 3427927.95 5793243.95,3427928 5793243.95,3427928.05 5793243.95,3427928.05 5793243.9,3427928 5793243.9)),((3427927.95 5793243.9,3427927.95 579324 3.85,3427927.9 5793243.85,3427927.85 5793243.85,3427927.85 5793243.9,3427927.9 5793243.9,3427927.9 5793243.95, 3427927.95 5793243.95,3427927.95 5793243.9)),((3427927.85 5793243.75,3427927.85 5793243.7,3427927.8 5793243.7,3427927.8 5793243.75 ,3427927.8 5793243.8,3427927.8 5793243.85,3427927.85 5793243.85,3427927.85 5793243.8,3427927.85 5793243.75)), ((3427928.05 5793243.75,3427928.05 5793243.7,3427928 5793243.7,3427927.95 5793243.7,3427927.95 5793243.75,3427927.95 5793243.8,3427 927.95 5793243.85,3427928 5793243.85,3427928 5793243.8,3427928.05 5793243.8, 3427928.05 5793243.75)),((3427927.95 5793243.75,3427927.95 5793243.7,3427927.9 5793243.7,3427927.85 5793243.7, 3427927.85 5793243.75,3427927.85 5793243.8,3427927.85 5793243.85,3427927.9 5793243.85, 3427927.95 5793243.85,3427927.95 5793243.8,3427927.95 5793243.75)))
--- Checking all the pixels of a large raster tile can take a long time. --- You can dramatically improve speed at some lose of precision by orders of magnitude -- by sampling pixels using the step optional parameter of generate_series. -- This next example does the same as previous but by checking 1 for every 4 (2x2) pixels and putting in the last checked -- putting in the checked pixel as the value for subsequent 4 SELECT ST_AsText(ST_Union(pixpolyg)) As shadow FROM (SELECT ST_Translate(ST_MakeEnvelope( ST_UpperLeftX(rast), ST_UpperLeftY(rast), ST_UpperLeftX(rast) + ST_ScaleX(rast)*2, ST_UpperLeftY(rast) + ST_ScaleY(rast)*2, 0 ), ST_ScaleX(rast)*x, ST_ScaleY(rast)*y ) As pixpolyg, ST_Value(rast, 2, x, y) As b2val FROM dummy_rast CROSS JOIN generate_series(1,1000,2) As x CROSS JOIN generate_series(1,1000,2) As y WHERE rid = 2 AND x <= ST_Width(rast) AND y <= ST_Height(rast) ) As foo WHERE ST_Intersects( pixpolyg, ST_GeomFromText('POLYGON((3427928 5793244,3427927.75 5793243.75,3427928 5793243.75,3427928 5793244))',0) ) AND b2val != 254; shadow ------------------------------------------------------------------------------------ MULTIPOLYGON(((3427927.9 5793243.85,3427927.8 5793243.85,3427927.8 5793243.95, 3427927.9 5793243.95,3427928 5793243.95,3427928.1 5793243.95,3427928.1 5793243.85,3427928 5793243.85,3427927.9 5793243.85)), ((3427927.9 5793243.65,3427927.8 5793243.65,3427927.8 5793243.75,3427927.8 5793243.85,3427927.9 5793243.85, 3427928 5793243.85,3427928 5793243.75,3427928.1 5793243.75,3427928.1 5793243.65,3427928 5793243.65,3427927.9 5793243.65)))
ST_NearestValue — Returns the nearest non-NODATA
value of a given band's pixel specified by a columnx and rowy or a geometric point expressed in the same spatial reference coordinate system as the raster.
double precision ST_NearestValue(
raster rast, integer bandnum, geometry pt, boolean exclude_nodata_value=true)
;
double precision ST_NearestValue(
raster rast, geometry pt, boolean exclude_nodata_value=true)
;
double precision ST_NearestValue(
raster rast, integer bandnum, integer columnx, integer rowy, boolean exclude_nodata_value=true)
;
double precision ST_NearestValue(
raster rast, integer columnx, integer rowy, boolean exclude_nodata_value=true)
;
Returns the nearest non-NODATA
value of a given band in a given columnx, rowy pixel or at a specific geometric point. If the columnx, rowy pixel or the pixel at the specified geometric point is NODATA
, the function will find the nearest pixel to the columnx, rowy pixel or geometric point whose value is not NODATA
.
Band numbers start at 1 and bandnum
is assumed to be 1 if not specified. If exclude_nodata_value
is set to false, then all pixels include nodata
pixels are considered to intersect and return value. If exclude_nodata_value
is not passed in then reads it from metadata of raster.
Availability: 2.1.0
ST_NearestValue is a drop-in replacement for ST_Value. |
-- pixel 2x2 has value SELECT ST_Value(rast, 2, 2) AS value, ST_NearestValue(rast, 2, 2) AS nearestvalue FROM ( SELECT ST_SetValue( ST_SetValue( ST_SetValue( ST_SetValue( ST_SetValue( ST_AddBand( ST_MakeEmptyRaster(5, 5, -2, 2, 1, -1, 0, 0, 0), '8BUI'::text, 1, 0 ), 1, 1, 0. ), 2, 3, 0. ), 3, 5, 0. ), 4, 2, 0. ), 5, 4, 0. ) AS rast ) AS foo value | nearestvalue -------+-------------- 1 | 1
-- pixel 2x3 is NODATA SELECT ST_Value(rast, 2, 3) AS value, ST_NearestValue(rast, 2, 3) AS nearestvalue FROM ( SELECT ST_SetValue( ST_SetValue( ST_SetValue( ST_SetValue( ST_SetValue( ST_AddBand( ST_MakeEmptyRaster(5, 5, -2, 2, 1, -1, 0, 0, 0), '8BUI'::text, 1, 0 ), 1, 1, 0. ), 2, 3, 0. ), 3, 5, 0. ), 4, 2, 0. ), 5, 4, 0. ) AS rast ) AS foo value | nearestvalue -------+-------------- | 1
ST_Neighborhood — Returns a 2-D double precision array of the non-NODATA
values around a given band's pixel specified by either a columnX and rowY or a geometric point expressed in the same spatial reference coordinate system as the raster.
double precision[][] ST_Neighborhood(
raster rast, integer bandnum, integer columnX, integer rowY, integer distanceX, integer distanceY, boolean exclude_nodata_value=true)
;
double precision[][] ST_Neighborhood(
raster rast, integer columnX, integer rowY, integer distanceX, integer distanceY, boolean exclude_nodata_value=true)
;
double precision[][] ST_Neighborhood(
raster rast, integer bandnum, geometry pt, integer distanceX, integer distanceY, boolean exclude_nodata_value=true)
;
double precision[][] ST_Neighborhood(
raster rast, geometry pt, integer distanceX, integer distanceY, boolean exclude_nodata_value=true)
;
Returns a 2-D double precision array of the non-NODATA
values around a given band's pixel specified by either a columnX and rowY or a geometric point expressed in the same spatial reference coordinate system as the raster. The distanceX
and distanceY
parameters define the number of pixels around the specified pixel in the X and Y axes, e.g. I want all values within 3 pixel distance along the X axis and 2 pixel distance along the Y axis around my pixel of interest. The center value of the 2-D array will be the value at the pixel specified by the columnX and rowY or the geometric point.
Band numbers start at 1 and bandnum
is assumed to be 1 if not specified. If exclude_nodata_value
is set to false, then all pixels include nodata
pixels are considered to intersect and return value. If exclude_nodata_value
is not passed in then reads it from metadata of raster.
The number of elements along each axis of the returning 2-D array is 2 * ( |
The 2-D array output can be passed to any of the raster processing builtin functions, e.g. ST_Min4ma, ST_Sum4ma, ST_Mean4ma. |
Availability: 2.1.0
-- un pixel 2x2 ha valore SELECT ST_Neighborhood(rast, 2, 2, 1, 1) FROM ( SELECT ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(5, 5, -2, 2, 1, -1, 0, 0, 0), '8BUI'::text, 1, 0 ), 1, 1, 1, ARRAY[ [0, 1, 1, 1, 1], [1, 1, 1, 0, 1], [1, 0, 1, 1, 1], [1, 1, 1, 1, 0], [1, 1, 0, 1, 1] ]::double precision[], 1 ) AS rast ) AS foo st_neighborhood --------------------------------- {{NULL,1,1},{1,1,NULL},{1,1,1}}
-- pixel 2x3 is NODATA SELECT ST_Neighborhood(rast, 2, 3, 1, 1) FROM ( SELECT ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(5, 5, -2, 2, 1, -1, 0, 0, 0), '8BUI'::text, 1, 0 ), 1, 1, 1, ARRAY[ [0, 1, 1, 1, 1], [1, 1, 1, 0, 1], [1, 0, 1, 1, 1], [1, 1, 1, 1, 0], [1, 1, 0, 1, 1] ]::double precision[], 1 ) AS rast ) AS foo st_neighborhood ------------------------------ {{1,1,1},{1,NULL,1},{1,1,1}}
-- pixel 3x3 has value -- exclude_nodata_value = FALSE SELECT ST_Neighborhood(rast, 3, 3, 1, 1, false) FROM ( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(5, 5, -2, 2, 1, -1, 0, 0, 0), '8BUI'::text, 1, 0 ), 1, 1, 1, ARRAY[ [0, 1, 1, 1, 1], [1, 1, 1, 0, 1], [1, 0, 1, 1, 1], [1, 1, 1, 1, 0], [1, 1, 0, 1, 1] ]::double precision[], 1 ) AS rast ) AS foo st_neighborhood --------------------------- {{1,0,1},{1,1,1},{0,1,1}}
ST_NearestValue, ST_Min4ma, ST_Max4ma, ST_Sum4ma, ST_Mean4ma, ST_Range4ma, ST_Distinct4ma, ST_StdDev4ma
ST_SetValue — Returns modified raster resulting from setting the value of a given band in a given columnx, rowy pixel or the pixels that intersect a particular geometry. Band numbers start at 1 and assumed to be 1 if not specified.
raster ST_SetValue(
raster rast, integer bandnum, geometry geom, double precision newvalue)
;
raster ST_SetValue(
raster rast, geometry geom, double precision newvalue)
;
raster ST_SetValue(
raster rast, integer bandnum, integer columnx, integer rowy, double precision newvalue)
;
raster ST_SetValue(
raster rast, integer columnx, integer rowy, double precision newvalue)
;
Returns modified raster resulting from setting the specified pixels' values to new value for the designed band given the raster's row and column or a geometry. If no band is specified, then band 1 is assumed.
Enhanced: 2.1.0 Geometry variant of ST_SetValue() now supports any geometry type, not just point. The geometry variant is a wrapper around the geomval[] variant of ST_SetValues()
-- Geometry example SELECT (foo.geomval).val, ST_AsText(ST_Union((foo.geomval).geom)) FROM (SELECT ST_DumpAsPolygons( ST_SetValue(rast,1, ST_Point(3427927.75, 5793243.95), 50) ) As geomval FROM dummy_rast where rid = 2) As foo WHERE (foo.geomval).val < 250 GROUP BY (foo.geomval).val; val | st_astext -----+------------------------------------------------------------------- 50 | POLYGON((3427927.75 5793244,3427927.75 5793243.95,3427927.8 579324 ... 249 | POLYGON((3427927.95 5793243.95,3427927.95 5793243.85,3427928 57932 ...
-- Store the changed raster -- UPDATE dummy_rast SET rast = ST_SetValue(rast,1, ST_Point(3427927.75, 5793243.95),100) WHERE rid = 2 ;
ST_SetValues — Returns modified raster resulting from setting the values of a given band.
raster ST_SetValues(
raster rast, integer nband, integer columnx, integer rowy, double precision[][] newvalueset, boolean[][] noset=NULL, boolean keepnodata=FALSE)
;
raster ST_SetValues(
raster rast, integer nband, integer columnx, integer rowy, double precision[][] newvalueset, double precision nosetvalue, boolean keepnodata=FALSE)
;
raster ST_SetValues(
raster rast, integer nband, integer columnx, integer rowy, integer width, integer height, double precision newvalue, boolean keepnodata=FALSE)
;
raster ST_SetValues(
raster rast, integer columnx, integer rowy, integer width, integer height, double precision newvalue, boolean keepnodata=FALSE)
;
raster ST_SetValues(
raster rast, integer nband, geomval[] geomvalset, boolean keepnodata=FALSE)
;
Returns modified raster resulting from setting specified pixels to new value(s) for the designated band.
If keepnodata
is TRUE, those pixels whose values are NODATA will not be set with the corresponding value in newvalueset
.
For Variant 1, the specific pixels to be set are determined by the columnx
, rowy
pixel coordinates and the dimensions of the newvalueset
array. noset
can be used to prevent pixels with values present in newvalueset
from being set (due to PostgreSQL not permitting ragged/jagged arrays). See example Variant 1.
Variant 2 is like Variant 1 but with a simple double precision nosetvalue
instead of a boolean noset
array. Elements in newvalueset
with the nosetvalue
value with be skipped. See example Variant 2.
For Variant 3, the specific pixels to be set are determined by the columnx
, rowy
pixel coordinates, width
and height
. See example Variant 3.
Variant 4 is the same as Variant 3 with the exception that it assumes that the first band's pixels of rast
will be set.
For Variant 5, an array of geomval is used to determine the specific pixels to be set. If all the geometries in the array are of type POINT or MULTIPOINT, the function uses a shortcut where the longitude and latitude of each point is used to set a pixel directly. Otherwise, the geometries are converted to rasters and then iterated through in one pass. See example Variant 5.
Availability: 2.1.0
/* The ST_SetValues() does the following... + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 1 | 1 | + - + - + - + + - + - + - + | 1 | 1 | 1 | => | 1 | 9 | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 2, 2, ARRAY[[9, 9], [9, 9]]::double precision[][] ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 1 1 | 2 | 1 1 | 3 | 1 2 | 1 | 1 2 | 2 | 9 2 | 3 | 9 3 | 1 | 1 3 | 2 | 9 3 | 3 | 9
/* The ST_SetValues() does the following... + - + - + - + + - + - + - + | 1 | 1 | 1 | | 9 | 9 | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | => | 9 | | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 9 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 1, 1, ARRAY[[9, 9, 9], [9, NULL, 9], [9, 9, 9]]::double precision[][] ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 9 1 | 2 | 9 1 | 3 | 9 2 | 1 | 9 2 | 2 | 2 | 3 | 9 3 | 1 | 9 3 | 2 | 9 3 | 3 | 9
/* The ST_SetValues() does the following... + - + - + - + + - + - + - + | 1 | 1 | 1 | | 9 | 9 | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | => | 1 | | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 9 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 1, 1, ARRAY[[9, 9, 9], [9, NULL, 9], [9, 9, 9]]::double precision[][], ARRAY[[false], [true]]::boolean[][] ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 9 1 | 2 | 1 1 | 3 | 9 2 | 1 | 9 2 | 2 | 2 | 3 | 9 3 | 1 | 9 3 | 2 | 9 3 | 3 | 9
/* The ST_SetValues() does the following... + - + - + - + + - + - + - + | | 1 | 1 | | | 9 | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | => | 1 | | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 9 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_SetValue( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 1, 1, NULL ), 1, 1, 1, ARRAY[[9, 9, 9], [9, NULL, 9], [9, 9, 9]]::double precision[][], ARRAY[[false], [true]]::boolean[][], TRUE ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 1 | 2 | 1 1 | 3 | 9 2 | 1 | 9 2 | 2 | 2 | 3 | 9 3 | 1 | 9 3 | 2 | 9 3 | 3 | 9
/* The ST_SetValues() does the following... + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 1 | 1 | + - + - + - + + - + - + - + | 1 | 1 | 1 | => | 1 | 9 | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 1, 1, ARRAY[[-1, -1, -1], [-1, 9, 9], [-1, 9, 9]]::double precision[][], -1 ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 1 1 | 2 | 1 1 | 3 | 1 2 | 1 | 1 2 | 2 | 9 2 | 3 | 9 3 | 1 | 1 3 | 2 | 9 3 | 3 | 9
/* This example is like the previous one. Instead of nosetvalue = -1, nosetvalue = NULL The ST_SetValues() does the following... + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 1 | 1 | + - + - + - + + - + - + - + | 1 | 1 | 1 | => | 1 | 9 | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 1, 1, ARRAY[[NULL, NULL, NULL], [NULL, 9, 9], [NULL, 9, 9]]::double precision[][], NULL::double precision ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 1 1 | 2 | 1 1 | 3 | 1 2 | 1 | 1 2 | 2 | 9 2 | 3 | 9 3 | 1 | 1 3 | 2 | 9 3 | 3 | 9
/* The ST_SetValues() does the following... + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 1 | 1 | + - + - + - + + - + - + - + | 1 | 1 | 1 | => | 1 | 9 | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 2, 2, 2, 2, 9 ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 1 1 | 2 | 1 1 | 3 | 1 2 | 1 | 1 2 | 2 | 9 2 | 3 | 9 3 | 1 | 1 3 | 2 | 9 3 | 3 | 9
/* The ST_SetValues() does the following... + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 1 | 1 | + - + - + - + + - + - + - + | 1 | | 1 | => | 1 | | 9 | + - + - + - + + - + - + - + | 1 | 1 | 1 | | 1 | 9 | 9 | + - + - + - + + - + - + - + */ SELECT (poly).x, (poly).y, (poly).val FROM ( SELECT ST_PixelAsPolygons( ST_SetValues( ST_SetValue( ST_AddBand( ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0 ), 1, 2, 2, NULL ), 1, 2, 2, 2, 2, 9, TRUE ) ) AS poly ) foo ORDER BY 1, 2; x | y | val ---+---+----- 1 | 1 | 1 1 | 2 | 1 1 | 3 | 1 2 | 1 | 1 2 | 2 | 2 | 3 | 9 3 | 1 | 1 3 | 2 | 9 3 | 3 | 9
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_MakeEmptyRaster(5, 5, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 0, 0) AS rast ), bar AS ( SELECT 1 AS gid, 'SRID=0;POINT(2.5 -2.5)'::geometry geom UNION ALL SELECT 2 AS gid, 'SRID=0;POLYGON((1 -1, 4 -1, 4 -4, 1 -4, 1 -1))'::geometry geom UNION ALL SELECT 3 AS gid, 'SRID=0;POLYGON((0 0, 5 0, 5 -1, 1 -1, 1 -4, 0 -4, 0 0))'::geometry geom UNION ALL SELECT 4 AS gid, 'SRID=0;MULTIPOINT(0 0, 4 4, 4 -4)'::geometry ) SELECT rid, gid, ST_DumpValues(ST_SetValue(rast, 1, geom, gid)) FROM foo t1 CROSS JOIN bar t2 ORDER BY rid, gid; rid | gid | st_dumpvalues -----+-----+--------------------------------------------------------------------------------------------------------------------------------------------- 1 | 1 | (1,"{{NULL,NULL,NULL,NULL,NULL},{NULL,NULL,NULL,NULL,NULL},{NULL,NULL,1,NULL,NULL},{NULL,NULL,NULL,NULL,NULL},{NULL,NULL,NULL,NULL,NULL}}") 1 | 2 | (1,"{{NULL,NULL,NULL,NULL,NULL},{NULL,2,2,2,NULL},{NULL,2,2,2,NULL},{NULL,2,2,2,NULL},{NULL,NULL,NULL,NULL,NULL}}") 1 | 3 | (1,"{{3,3,3,3,3},{3,NULL,NULL,NULL,NULL},{3,NULL,NULL,NULL,NULL},{3,NULL,NULL,NULL,NULL},{NULL,NULL,NULL,NULL,NULL}}") 1 | 4 | (1,"{{4,NULL,NULL,NULL,NULL},{NULL,NULL,NULL,NULL,NULL},{NULL,NULL,NULL,NULL,NULL},{NULL,NULL,NULL,NULL,NULL},{NULL,NULL,NULL,NULL,4}}") (4 rows)
The following shows that geomvals later in the array can overwrite prior geomvals
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_MakeEmptyRaster(5, 5, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 0, 0) AS rast ), bar AS ( SELECT 1 AS gid, 'SRID=0;POINT(2.5 -2.5)'::geometry geom UNION ALL SELECT 2 AS gid, 'SRID=0;POLYGON((1 -1, 4 -1, 4 -4, 1 -4, 1 -1))'::geometry geom UNION ALL SELECT 3 AS gid, 'SRID=0;POLYGON((0 0, 5 0, 5 -1, 1 -1, 1 -4, 0 -4, 0 0))'::geometry geom UNION ALL SELECT 4 AS gid, 'SRID=0;MULTIPOINT(0 0, 4 4, 4 -4)'::geometry ) SELECT t1.rid, t2.gid, t3.gid, ST_DumpValues(ST_SetValues(rast, 1, ARRAY[ROW(t2.geom, t2.gid), ROW(t3.geom, t3.gid)]::geomval[])) FROM foo t1 CROSS JOIN bar t2 CROSS JOIN bar t3 WHERE t2.gid = 1 AND t3.gid = 2 ORDER BY t1.rid, t2.gid, t3.gid; rid | gid | gid | st_dumpvalues -----+-----+-----+--------------------------------------------------------------------------------------------------------------------- 1 | 1 | 2 | (1,"{{NULL,NULL,NULL,NULL,NULL},{NULL,2,2,2,NULL},{NULL,2,2,2,NULL},{NULL,2,2,2,NULL},{NULL,NULL,NULL,NULL,NULL}}") (1 row)
This example is the opposite of the prior example
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_MakeEmptyRaster(5, 5, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 0, 0) AS rast ), bar AS ( SELECT 1 AS gid, 'SRID=0;POINT(2.5 -2.5)'::geometry geom UNION ALL SELECT 2 AS gid, 'SRID=0;POLYGON((1 -1, 4 -1, 4 -4, 1 -4, 1 -1))'::geometry geom UNION ALL SELECT 3 AS gid, 'SRID=0;POLYGON((0 0, 5 0, 5 -1, 1 -1, 1 -4, 0 -4, 0 0))'::geometry geom UNION ALL SELECT 4 AS gid, 'SRID=0;MULTIPOINT(0 0, 4 4, 4 -4)'::geometry ) SELECT t1.rid, t2.gid, t3.gid, ST_DumpValues(ST_SetValues(rast, 1, ARRAY[ROW(t2.geom, t2.gid), ROW(t3.geom, t3.gid)]::geomval[])) FROM foo t1 CROSS JOIN bar t2 CROSS JOIN bar t3 WHERE t2.gid = 2 AND t3.gid = 1 ORDER BY t1.rid, t2.gid, t3.gid; rid | gid | gid | st_dumpvalues -----+-----+-----+--------------------------------------------------------------------------------------------------------------------- 1 | 2 | 1 | (1,"{{NULL,NULL,NULL,NULL,NULL},{NULL,2,2,2,NULL},{NULL,2,1,2,NULL},{NULL,2,2,2,NULL},{NULL,NULL,NULL,NULL,NULL}}") (1 row)
ST_DumpValues — Get the values of the specified band as a 2-dimension array.
setof record ST_DumpValues(
raster rast , integer[] nband=NULL , boolean exclude_nodata_value=true )
;
double precision[][] ST_DumpValues(
raster rast , integer nband , boolean exclude_nodata_value=true )
;
Get the values of the specified band as a 2-dimension array (first index is row, second is column). If nband
is NULL or not provided, all raster bands are processed.
Availability: 2.1.0
WITH foo AS ( SELECT ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0), 2, '32BF', 3, -9999), 3, '16BSI', 0, 0) AS rast ) SELECT (ST_DumpValues(rast)).* FROM foo; nband | valarray -------+------------------------------------------------------ 1 | {{1,1,1},{1,1,1},{1,1,1}} 2 | {{3,3,3},{3,3,3},{3,3,3}} 3 | {{NULL,NULL,NULL},{NULL,NULL,NULL},{NULL,NULL,NULL}} (3 rows)
WITH foo AS ( SELECT ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0), 2, '32BF', 3, -9999), 3, '16BSI', 0, 0) AS rast ) SELECT (ST_DumpValues(rast, ARRAY[3, 1])).* FROM foo; nband | valarray -------+------------------------------------------------------ 3 | {{NULL,NULL,NULL},{NULL,NULL,NULL},{NULL,NULL,NULL}} 1 | {{1,1,1},{1,1,1},{1,1,1}} (2 rows)
WITH foo AS ( SELECT ST_SetValue(ST_AddBand(ST_MakeEmptyRaster(3, 3, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 1, 0), 1, 2, 5) AS rast ) SELECT (ST_DumpValues(rast, 1))[2][1] FROM foo; st_dumpvalues --------------- 5 (1 row)
ST_PixelOfValue — Get the columnx, rowy coordinates of the pixel whose value equals the search value.
setof record ST_PixelOfValue(
raster rast , integer nband , double precision[] search , boolean exclude_nodata_value=true )
;
setof record ST_PixelOfValue(
raster rast , double precision[] search , boolean exclude_nodata_value=true )
;
setof record ST_PixelOfValue(
raster rast , integer nband , double precision search , boolean exclude_nodata_value=true )
;
setof record ST_PixelOfValue(
raster rast , double precision search , boolean exclude_nodata_value=true )
;
Get the columnx, rowy coordinates of the pixel whose value equals the search value. If no band is specified, then band 1 is assumed.
Availability: 2.1.0
SELECT (pixels).* FROM ( SELECT ST_PixelOfValue( ST_SetValue( ST_SetValue( ST_SetValue( ST_SetValue( ST_SetValue( ST_AddBand( ST_MakeEmptyRaster(5, 5, -2, 2, 1, -1, 0, 0, 0), '8BUI'::text, 1, 0 ), 1, 1, 0 ), 2, 3, 0 ), 3, 5, 0 ), 4, 2, 0 ), 5, 4, 255 ) , 1, ARRAY[1, 255]) AS pixels ) AS foo val | x | y -----+---+--- 1 | 1 | 2 1 | 1 | 3 1 | 1 | 4 1 | 1 | 5 1 | 2 | 1 1 | 2 | 2 1 | 2 | 4 1 | 2 | 5 1 | 3 | 1 1 | 3 | 2 1 | 3 | 3 1 | 3 | 4 1 | 4 | 1 1 | 4 | 3 1 | 4 | 4 1 | 4 | 5 1 | 5 | 1 1 | 5 | 2 1 | 5 | 3 255 | 5 | 4 1 | 5 | 5
ST_SetGeoReference — Set Georeference 6 georeference parameters in a single call. Numbers should be separated by white space. Accepts inputs in GDAL or ESRI format. Default is GDAL.
raster ST_SetGeoReference(
raster rast, text georefcoords, text format=GDAL)
;
raster ST_SetGeoReference(
raster rast, double precision upperleftx, double precision upperlefty, double precision scalex, double precision scaley, double precision skewx, double precision skewy)
;
Set Georeference 6 georeference parameters in a single call. Accepts inputs in 'GDAL' or 'ESRI' format. Default is GDAL. If 6 coordinates are not provided will return null.
Difference between format representations is as follows:
GDAL
:
scalex skewy skewx scaley upperleftx upperlefty
ESRI
:
scalex skewy skewx scaley upperleftx + scalex*0.5 upperlefty + scaley*0.5
If the raster has out-db bands, changing the georeference may result in incorrect access of the band's externally stored data. |
Enhanced: 2.1.0 Addition of ST_SetGeoReference(raster, double precision, ...) variant
WITH foo AS ( SELECT ST_MakeEmptyRaster(5, 5, 0, 0, 1, -1, 0, 0, 0) AS rast ) SELECT 0 AS rid, (ST_Metadata(rast)).* FROM foo UNION ALL SELECT 1, (ST_Metadata(ST_SetGeoReference(rast, '10 0 0 -10 0.1 0.1', 'GDAL'))).* FROM foo UNION ALL SELECT 2, (ST_Metadata(ST_SetGeoReference(rast, '10 0 0 -10 5.1 -4.9', 'ESRI'))).* FROM foo UNION ALL SELECT 3, (ST_Metadata(ST_SetGeoReference(rast, 1, 1, 10, -10, 0.001, 0.001))).* FROM foo rid | upperleftx | upperlefty | width | height | scalex | scaley | skewx | skewy | srid | numbands -----+--------------------+--------------------+-------+--------+--------+--------+-------+-------+------+---------- 0 | 0 | 0 | 5 | 5 | 1 | -1 | 0 | 0 | 0 | 0 1 | 0.1 | 0.1 | 5 | 5 | 10 | -10 | 0 | 0 | 0 | 0 2 | 0.0999999999999996 | 0.0999999999999996 | 5 | 5 | 10 | -10 | 0 | 0 | 0 | 0 3 | 1 | 1 | 5 | 5 | 10 | -10 | 0.001 | 0.001 | 0 | 0
ST_SetRotation — Set the rotation of the raster in radian.
float8 ST_SetRotation(
raster rast, float8 rotation)
;
Uniformly rotate the raster. Rotation is in radian. Refer to World File for more details.
SELECT ST_ScaleX(rast1), ST_ScaleY(rast1), ST_SkewX(rast1), ST_SkewY(rast1), ST_ScaleX(rast2), ST_ScaleY(rast2), ST_SkewX(rast2), ST_SkewY(rast2) FROM ( SELECT ST_SetRotation(rast, 15) AS rast1, rast as rast2 FROM dummy_rast ) AS foo; st_scalex | st_scaley | st_skewx | st_skewy | st_scalex | st_scaley | st_skewx | st_skewy ---------------------+---------------------+--------------------+--------------------+-----------+-----------+----------+---------- -1.51937582571764 | -2.27906373857646 | 1.95086352047135 | 1.30057568031423 | 2 | 3 | 0 | 0 -0.0379843956429411 | -0.0379843956429411 | 0.0325143920078558 | 0.0325143920078558 | 0.05 | -0.05 | 0 | 0
ST_SetScale — Sets the X and Y size of pixels in units of coordinate reference system. Number units/pixel width/height.
raster ST_SetScale(
raster rast, float8 xy)
;
raster ST_SetScale(
raster rast, float8 x, float8 y)
;
Sets the X and Y size of pixels in units of coordinate reference system. Number units/pixel width/height. If only one unit passed in, assumed X and Y are the same number.
ST_SetScale is different from ST_Rescale in that ST_SetScale do not resample the raster to match the raster extent. It only changes the metadata (or georeference) of the raster to correct an originally mis-specified scaling. ST_Rescale results in a raster having different width and height computed to fit the geographic extent of the input raster. ST_SetScale do not modify the width, nor the height of the raster. |
Changed: 2.0.0 In WKTRaster versions this was called ST_SetPixelSize. This was changed in 2.0.0.
UPDATE dummy_rast SET rast = ST_SetScale(rast, 1.5) WHERE rid = 2; SELECT ST_ScaleX(rast) As pixx, ST_ScaleY(rast) As pixy, Box3D(rast) As newbox FROM dummy_rast WHERE rid = 2; pixx | pixy | newbox ------+------+---------------------------------------------- 1.5 | 1.5 | BOX(3427927.75 5793244 0, 3427935.25 5793251.5 0)
UPDATE dummy_rast SET rast = ST_SetScale(rast, 1.5, 0.55) WHERE rid = 2; SELECT ST_ScaleX(rast) As pixx, ST_ScaleY(rast) As pixy, Box3D(rast) As newbox FROM dummy_rast WHERE rid = 2; pixx | pixy | newbox ------+------+-------------------------------------------- 1.5 | 0.55 | BOX(3427927.75 5793244 0,3427935.25 5793247 0)
ST_SetSkew — Sets the georeference X and Y skew (or rotation parameter). If only one is passed in, sets X and Y to the same value.
raster ST_SetSkew(
raster rast, float8 skewxy)
;
raster ST_SetSkew(
raster rast, float8 skewx, float8 skewy)
;
Sets the georeference X and Y skew (or rotation parameter). If only one is passed in, sets X and Y to the same value. Refer to World File for more details.
-- Example 1 UPDATE dummy_rast SET rast = ST_SetSkew(rast,1,2) WHERE rid = 1; SELECT rid, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy, ST_GeoReference(rast) as georef FROM dummy_rast WHERE rid = 1; rid | skewx | skewy | georef ----+-------+-------+-------------- 1 | 1 | 2 | 2.0000000000 : 2.0000000000 : 1.0000000000 : 3.0000000000 : 0.5000000000 : 0.5000000000
-- Example 2 set both to same number: UPDATE dummy_rast SET rast = ST_SetSkew(rast,0) WHERE rid = 1; SELECT rid, ST_SkewX(rast) As skewx, ST_SkewY(rast) As skewy, ST_GeoReference(rast) as georef FROM dummy_rast WHERE rid = 1; rid | skewx | skewy | georef -----+-------+-------+-------------- 1 | 0 | 0 | 2.0000000000 : 0.0000000000 : 0.0000000000 : 3.0000000000 : 0.5000000000 : 0.5000000000
ST_SetSRID — Sets the SRID of a raster to a particular integer srid defined in the spatial_ref_sys table.
raster ST_SetSRID(
raster rast, integer srid)
;
ST_SetUpperLeft — Sets the value of the upper left corner of the pixel to projected X and Y coordinates.
raster ST_SetUpperLeft(
raster rast, double precision x, double precision y)
;
ST_Resample — Resample a raster using a specified resampling algorithm, new dimensions, an arbitrary grid corner and a set of raster georeferencing attributes defined or borrowed from another raster.
raster ST_Resample(
raster rast, integer width, integer height, double precision gridx=NULL, double precision gridy=NULL, double precision skewx=0, double precision skewy=0, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
raster ST_Resample(
raster rast, double precision scalex=0, double precision scaley=0, double precision gridx=NULL, double precision gridy=NULL, double precision skewx=0, double precision skewy=0, text algorithm=NearestNeighbor, double precision maxerr=0.125)
;
raster ST_Resample(
raster rast, raster ref, text algorithm=NearestNeighbour, double precision maxerr=0.125, boolean usescale=true)
;
raster ST_Resample(
raster rast, raster ref, boolean usescale, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
Resample a raster using a specified resampling algorithm, new dimensions (width & height), a grid corner (gridx & gridy) and a set of raster georeferencing attributes (scalex, scaley, skewx & skewy) defined or borrowed from another raster. If using a reference raster, the two rasters must have the same SRID.
New pixel values are computed using the NearestNeighbor (English or American spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. Default is NearestNeighbor which is the fastest but produce the worst interpolation.
A maxerror percent of 0.125 is used if no maxerr
is specified.
Refer to: GDAL Warp resampling methods for more details. |
Availability: 2.0.0 Requires GDAL 1.6.1+
Changed: 2.1.0 Parameter srid removed. Variants with a reference raster no longer applies the reference raster's SRID. Use ST_Transform() to reproject raster. Works on rasters with no SRID.
SELECT ST_Width(orig) AS orig_width, ST_Width(reduce_100) AS new_width FROM ( SELECT rast AS orig, ST_Resample(rast,100,100) AS reduce_100 FROM aerials.boston WHERE ST_Intersects(rast, ST_Transform( ST_MakeEnvelope(-71.128, 42.2392,-71.1277, 42.2397, 4326),26986) ) LIMIT 1 ) AS foo; orig_width | new_width ------------+------------- 200 | 100
ST_Rescale — Resample a raster by adjusting only its scale (or pixel size). New pixel values are computed using the NearestNeighbor (english or american spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. Default is NearestNeighbor.
raster ST_Rescale(
raster rast, double precision scalexy, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
raster ST_Rescale(
raster rast, double precision scalex, double precision scaley, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
Resample a raster by adjusting only its scale (or pixel size). New pixel values are computed using the NearestNeighbor (english or american spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. The default is NearestNeighbor which is the fastest but results in the worst interpolation.
scalex
and scaley
define the new pixel size. scaley must often be negative to get well oriented raster.
When the new scalex or scaley is not a divisor of the raster width or height, the extent of the resulting raster is expanded to encompass the extent of the provided raster. If you want to be sure to retain exact input extent see ST_Resize
A maxerror percent of 0.125 is used if no maxerr
is specified.
Refer to: GDAL Warp resampling methods for more details. |
ST_Rescale is different from ST_SetScale in that ST_SetScale do not resample the raster to match the raster extent. ST_SetScale only changes the metadata (or georeference) of the raster to correct an originally mis-specified scaling. ST_Rescale results in a raster having different width and height computed to fit the geographic extent of the input raster. ST_SetScale do not modify the width, nor the height of the raster. |
Availability: 2.0.0 Requires GDAL 1.6.1+
Changed: 2.1.0 Works on rasters with no SRID
A simple example rescaling a raster from a pixel size of 0.001 degree to a pixel size of 0.0015 degree.
-- the original raster pixel size SELECT ST_PixelWidth(ST_AddBand(ST_MakeEmptyRaster(100, 100, 0, 0, 0.001, -0.001, 0, 0, 4269), '8BUI'::text, 1, 0)) width width ---------- 0.001 -- the rescaled raster raster pixel size SELECT ST_PixelWidth(ST_Rescale(ST_AddBand(ST_MakeEmptyRaster(100, 100, 0, 0, 0.001, -0.001, 0, 0, 4269), '8BUI'::text, 1, 0), 0.0015)) width width ---------- 0.0015
ST_Reskew — Resample a raster by adjusting only its skew (or rotation parameters). New pixel values are computed using the NearestNeighbor (english or american spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. Default is NearestNeighbor.
raster ST_Reskew(
raster rast, double precision skewxy, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
raster ST_Reskew(
raster rast, double precision skewx, double precision skewy, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
Resample a raster by adjusting only its skew (or rotation parameters). New pixel values are computed using the NearestNeighbor (english or american spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. The default is NearestNeighbor which is the fastest but results in the worst interpolation.
skewx
and skewy
define the new skew.
The extent of the new raster will encompass the extent of the provided raster.
A maxerror percent of 0.125 if no maxerr
is specified.
Refer to: GDAL Warp resampling methods for more details. |
ST_Reskew is different from ST_SetSkew in that ST_SetSkew do not resample the raster to match the raster extent. ST_SetSkew only changes the metadata (or georeference) of the raster to correct an originally mis-specified skew. ST_Reskew results in a raster having different width and height computed to fit the geographic extent of the input raster. ST_SetSkew do not modify the width, nor the height of the raster. |
Availability: 2.0.0 Requires GDAL 1.6.1+
Changed: 2.1.0 Works on rasters with no SRID
A simple example reskewing a raster from a skew of 0.0 to a skew of 0.0015.
-- the original raster pixel size SELECT ST_Rotation(ST_AddBand(ST_MakeEmptyRaster(100, 100, 0, 0, 0.001, -0.001, 0, 0, 4269), '8BUI'::text, 1, 0)) -- the rescaled raster raster pixel size SELECT ST_Rotation(ST_Reskew(ST_AddBand(ST_MakeEmptyRaster(100, 100, 0, 0, 0.001, -0.001, 0, 0, 4269), '8BUI'::text, 1, 0), 0.0015))
ST_SnapToGrid — Resample a raster by snapping it to a grid. New pixel values are computed using the NearestNeighbor (english or american spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. Default is NearestNeighbor.
raster ST_SnapToGrid(
raster rast, double precision gridx, double precision gridy, text algorithm=NearestNeighbour, double precision maxerr=0.125, double precision scalex=DEFAULT 0, double precision scaley=DEFAULT 0)
;
raster ST_SnapToGrid(
raster rast, double precision gridx, double precision gridy, double precision scalex, double precision scaley, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
raster ST_SnapToGrid(
raster rast, double precision gridx, double precision gridy, double precision scalexy, text algorithm=NearestNeighbour, double precision maxerr=0.125)
;
Resample a raster by snapping it to a grid defined by an arbitrary pixel corner (gridx & gridy) and optionally a pixel size (scalex & scaley). New pixel values are computed using the NearestNeighbor (english or american spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. The default is NearestNeighbor which is the fastest but results in the worst interpolation.
gridx
and gridy
define any arbitrary pixel corner of the new grid. This is not necessarily the upper left corner of the new raster and it does not have to be inside or on the edge of the new raster extent.
You can optionnal define the pixel size of the new grid with scalex
and scaley
.
The extent of the new raster will encompass the extent of the provided raster.
A maxerror percent of 0.125 if no maxerr
is specified.
Refer to: GDAL Warp resampling methods for more details. |
Use ST_Resample if you need more control over the grid parameters. |
Availability: 2.0.0 Requires GDAL 1.6.1+
Changed: 2.1.0 Works on rasters with no SRID
A simple example snapping a raster to a slightly different grid.
-- the original raster pixel size SELECT ST_UpperLeftX(ST_AddBand(ST_MakeEmptyRaster(10, 10, 0, 0, 0.001, -0.001, 0, 0, 4269), '8BUI'::text, 1, 0)) -- the rescaled raster raster pixel size SELECT ST_UpperLeftX(ST_SnapToGrid(ST_AddBand(ST_MakeEmptyRaster(10, 10, 0, 0, 0.001, -0.001, 0, 0, 4269), '8BUI'::text, 1, 0), 0.0002, 0.0002))
ST_Resize — Resize a raster to a new width/height
raster ST_Resize(
raster rast, integer width, integer height, text algorithm=NearestNeighbor, double precision maxerr=0.125)
;
raster ST_Resize(
raster rast, double precision percentwidth, double precision percentheight, text algorithm=NearestNeighbor, double precision maxerr=0.125)
;
raster ST_Resize(
raster rast, text width, text height, text algorithm=NearestNeighbor, double precision maxerr=0.125)
;
Resize a raster to a new width/height. The new width/height can be specified in exact number of pixels or a percentage of the raster's width/height. The extent of the the new raster will be the same as the extent of the provided raster.
New pixel values are computed using the NearestNeighbor (english or american spelling), Bilinear, Cubic, CubicSpline or Lanczos resampling algorithm. The default is NearestNeighbor which is the fastest but results in the worst interpolation.
Variant 1 expects the actual width/height of the output raster.
Variant 2 expects decimal values between zero (0) and one (1) indicating the percentage of the input raster's width/height.
Variant 3 takes either the actual width/height of the output raster or a textual percentage ("20%") indicating the percentage of the input raster's width/height.
Availability: 2.1.0 Requires GDAL 1.6.1+
WITH foo AS( SELECT 1 AS rid, ST_Resize( ST_AddBand( ST_MakeEmptyRaster(1000, 1000, 0, 0, 1, -1, 0, 0, 0) , 1, '8BUI', 255, 0 ) , '50%', '500') AS rast UNION ALL SELECT 2 AS rid, ST_Resize( ST_AddBand( ST_MakeEmptyRaster(1000, 1000, 0, 0, 1, -1, 0, 0, 0) , 1, '8BUI', 255, 0 ) , 500, 100) AS rast UNION ALL SELECT 3 AS rid, ST_Resize( ST_AddBand( ST_MakeEmptyRaster(1000, 1000, 0, 0, 1, -1, 0, 0, 0) , 1, '8BUI', 255, 0 ) , 0.25, 0.9) AS rast ), bar AS ( SELECT rid, ST_Metadata(rast) AS meta, rast FROM foo ) SELECT rid, (meta).* FROM bar rid | upperleftx | upperlefty | width | height | scalex | scaley | skewx | skewy | srid | numbands -----+------------+------------+-------+--------+--------+--------+-------+-------+------+---------- 1 | 0 | 0 | 500 | 500 | 1 | -1 | 0 | 0 | 0 | 1 2 | 0 | 0 | 500 | 100 | 1 | -1 | 0 | 0 | 0 | 1 3 | 0 | 0 | 250 | 900 | 1 | -1 | 0 | 0 | 0 | 1 (3 rows)
ST_Transform — Reprojects a raster in a known spatial reference system to another known spatial reference system using specified resampling algorithm. Options are NearestNeighbor, Bilinear, Cubic, CubicSpline, Lanczos defaulting to NearestNeighbor.
raster ST_Transform(
raster rast, integer srid, text algorithm=NearestNeighbor, double precision maxerr=0.125, double precision scalex, double precision scaley)
;
raster ST_Transform(
raster rast, integer srid, double precision scalex, double precision scaley, text algorithm=NearestNeighbor, double precision maxerr=0.125)
;
raster ST_Transform(
raster rast, raster alignto, text algorithm=NearestNeighbor, double precision maxerr=0.125)
;
Reprojects a raster in a known spatial reference system to another known spatial reference system using specified pixel warping algorithm. Uses 'NearestNeighbor' if no algorithm is specified and maxerror percent of 0.125 if no maxerr is specified.
Algorithm options are: 'NearestNeighbor', 'Bilinear', 'Cubic', 'CubicSpline', and 'Lanczos'. Refer to: GDAL Warp resampling methods for more details.
ST_Transform is often confused with ST_SetSRID(). ST_Transform actually changes the coordinates of a raster (and resamples the pixel values) from one spatial reference system to another, while ST_SetSRID() simply changes the SRID identifier of the raster.
Unlike the other variants, Variant 3 requires a reference raster as alignto
. The transformed raster will be transformed to the spatial reference system (SRID) of the reference raster and be aligned (ST_SameAlignment = TRUE) to the reference raster.
If you find your transformation support is not working right, you may need to set the environment variable PROJSO to the .so or .dll projection library your PostGIS is using. This just needs to have the name of the file. So for example on windows, you would in Control Panel -> System -> Environment Variables add a system variable called |
Availability: 2.0.0 Requires GDAL 1.6.1+
Enhanced: 2.1.0 Addition of ST_Transform(rast, alignto) variant
SELECT ST_Width(mass_stm) As w_before, ST_Width(wgs_84) As w_after, ST_Height(mass_stm) As h_before, ST_Height(wgs_84) As h_after FROM ( SELECT rast As mass_stm, ST_Transform(rast,4326) As wgs_84 , ST_Transform(rast,4326, 'Bilinear') AS wgs_84_bilin FROM aerials.o_2_boston WHERE ST_Intersects(rast, ST_Transform(ST_MakeEnvelope(-71.128, 42.2392,-71.1277, 42.2397, 4326),26986) ) LIMIT 1) As foo; w_before | w_after | h_before | h_after ----------+---------+----------+--------- 200 | 228 | 200 | 170
The following shows the difference between using ST_Transform(raster, srid) and ST_Transform(raster, alignto)
WITH foo AS ( SELECT 0 AS rid, ST_AddBand(ST_MakeEmptyRaster(2, 2, -500000, 600000, 100, -100, 0, 0, 2163), 1, '16BUI', 1, 0) AS rast UNION ALL SELECT 1, ST_AddBand(ST_MakeEmptyRaster(2, 2, -499800, 600000, 100, -100, 0, 0, 2163), 1, '16BUI', 2, 0) AS rast UNION ALL SELECT 2, ST_AddBand(ST_MakeEmptyRaster(2, 2, -499600, 600000, 100, -100, 0, 0, 2163), 1, '16BUI', 3, 0) AS rast UNION ALL SELECT 3, ST_AddBand(ST_MakeEmptyRaster(2, 2, -500000, 599800, 100, -100, 0, 0, 2163), 1, '16BUI', 10, 0) AS rast UNION ALL SELECT 4, ST_AddBand(ST_MakeEmptyRaster(2, 2, -499800, 599800, 100, -100, 0, 0, 2163), 1, '16BUI', 20, 0) AS rast UNION ALL SELECT 5, ST_AddBand(ST_MakeEmptyRaster(2, 2, -499600, 599800, 100, -100, 0, 0, 2163), 1, '16BUI', 30, 0) AS rast UNION ALL SELECT 6, ST_AddBand(ST_MakeEmptyRaster(2, 2, -500000, 599600, 100, -100, 0, 0, 2163), 1, '16BUI', 100, 0) AS rast UNION ALL SELECT 7, ST_AddBand(ST_MakeEmptyRaster(2, 2, -499800, 599600, 100, -100, 0, 0, 2163), 1, '16BUI', 200, 0) AS rast UNION ALL SELECT 8, ST_AddBand(ST_MakeEmptyRaster(2, 2, -499600, 599600, 100, -100, 0, 0, 2163), 1, '16BUI', 300, 0) AS rast ), bar AS ( SELECT ST_Transform(rast, 4269) AS alignto FROM foo LIMIT 1 ), baz AS ( SELECT rid, rast, ST_Transform(rast, 4269) AS not_aligned, ST_Transform(rast, alignto) AS aligned FROM foo CROSS JOIN bar ) SELECT ST_SameAlignment(rast) AS rast, ST_SameAlignment(not_aligned) AS not_aligned, ST_SameAlignment(aligned) AS aligned FROM baz rast | not_aligned | aligned ------+-------------+--------- t | f | t
ST_SetBandNoDataValue — Sets the value for the given band that represents no data. Band 1 is assumed if no band is specified. To mark a band as having no nodata value, set the nodata value = NULL.
raster ST_SetBandNoDataValue(
raster rast, double precision nodatavalue)
;
raster ST_SetBandNoDataValue(
raster rast, integer band, double precision nodatavalue, boolean forcechecking=false)
;
Sets the value that represents no data for the band. Band 1 is assumed if not specified. This will affect results from ST_Polygon, ST_DumpAsPolygons, and the ST_PixelAs...() functions.
-- change just first band no data value UPDATE dummy_rast SET rast = ST_SetBandNoDataValue(rast,1, 254) WHERE rid = 2; -- change no data band value of bands 1,2,3 UPDATE dummy_rast SET rast = ST_SetBandNoDataValue( ST_SetBandNoDataValue( ST_SetBandNoDataValue( rast,1, 254) ,2,99), 3,108) WHERE rid = 2; -- wipe out the nodata value this will ensure all pixels are considered for all processing functions UPDATE dummy_rast SET rast = ST_SetBandNoDataValue(rast,1, NULL) WHERE rid = 2;
ST_SetBandIsNoData — Sets the isnodata flag of the band to TRUE.
raster ST_SetBandIsNoData(
raster rast, integer band=1)
;
Sets the isnodata flag for the band to true. Band 1 is assumed if not specified. This function should be called only when the flag is considered dirty. That is, when the result calling ST_BandIsNoData is different using TRUE as last argument and without using it
Disponibilità: 2.0
-- Create dummy table with one raster column create table dummy_rast (rid integer, rast raster); -- Add raster with two bands, one pixel/band. In the first band, nodatavalue = pixel value = 3. -- In the second band, nodatavalue = 13, pixel value = 4 insert into dummy_rast values(1, ( '01' -- little endian (uint8 ndr) || '0000' -- version (uint16 0) || '0200' -- nBands (uint16 0) || '17263529ED684A3F' -- scaleX (float64 0.000805965234044584) || 'F9253529ED684ABF' -- scaleY (float64 -0.00080596523404458) || '1C9F33CE69E352C0' -- ipX (float64 -75.5533328537098) || '718F0E9A27A44840' -- ipY (float64 49.2824585505576) || 'ED50EB853EC32B3F' -- skewX (float64 0.000211812383858707) || '7550EB853EC32B3F' -- skewY (float64 0.000211812383858704) || 'E6100000' -- SRID (int32 4326) || '0100' -- width (uint16 1) || '0100' -- height (uint16 1) || '4' -- hasnodatavalue set to true, isnodata value set to false (when it should be true) || '2' -- first band type (4BUI) || '03' -- novalue==3 || '03' -- pixel(0,0)==3 (same that nodata) || '0' -- hasnodatavalue set to false || '5' -- second band type (16BSI) || '0D00' -- novalue==13 || '0400' -- pixel(0,0)==4 )::raster ); select st_bandisnodata(rast, 1) from dummy_rast where rid = 1; -- Expected false select st_bandisnodata(rast, 1, TRUE) from dummy_rast where rid = 1; -- Expected true -- The isnodata flag is dirty. We are going to set it to true update dummy_rast set rast = st_setbandisnodata(rast, 1) where rid = 1; select st_bandisnodata(rast, 1) from dummy_rast where rid = 1; -- Expected true
ST_Count — Returns the number of pixels in a given band of a raster or raster coverage. If no band is specified defaults to band 1. If exclude_nodata_value is set to true, will only count pixels that are not equal to the nodata value.
bigint ST_Count(
raster rast, integer nband=1, boolean exclude_nodata_value=true)
;
bigint ST_Count(
raster rast, boolean exclude_nodata_value)
;
bigint ST_Count(
text rastertable, text rastercolumn, integer nband=1, boolean exclude_nodata_value=true)
;
bigint ST_Count(
text rastertable, text rastercolumn, boolean exclude_nodata_value)
;
Returns the number of pixels in a given band of a raster or raster coverage. If no band is specified nband
defaults to 1.
If |
Disponibilità: 2.0
The ST_Count(rastertable, rastercolumn, ...) variants are deprecated as of 2.2.0. Use ST_CountAgg instead. |
--example will count all pixels not 249 and one will count all pixels. -- SELECT rid, ST_Count(ST_SetBandNoDataValue(rast,249)) As exclude_nodata, ST_Count(ST_SetBandNoDataValue(rast,249),false) As include_nodata FROM dummy_rast WHERE rid=2; rid | exclude_nodata | include_nodata -----+----------------+---------------- 2 | 23 | 25
ST_CountAgg — Aggregate. Returns the number of pixels in a given band of a set of rasters. If no band is specified defaults to band 1. If exclude_nodata_value is set to true, will only count pixels that are not equal to the NODATA value.
bigint ST_CountAgg(
raster rast, integer nband, boolean exclude_nodata_value, double precision sample_percent)
;
bigint ST_CountAgg(
raster rast, integer nband, boolean exclude_nodata_value)
;
bigint ST_CountAgg(
raster rast, boolean exclude_nodata_value)
;
Returns the number of pixels in a given band of a set of rasters. If no band is specified nband
defaults to 1.
If exclude_nodata_value
is set to true, will only count pixels with value not equal to the NODATA
value of the raster. Set exclude_nodata_value
to false to get count all pixels
By default will sample all pixels. To get faster response, set sample_percent
to value between zero (0) and one (1)
Availability: 2.2.0
WITH foo AS ( SELECT rast.rast FROM ( SELECT ST_SetValue( ST_SetValue( ST_SetValue( ST_AddBand( ST_MakeEmptyRaster(10, 10, 10, 10, 2, 2, 0, 0,0) , 1, '64BF', 0, 0 ) , 1, 1, 1, -10 ) , 1, 5, 4, 0 ) , 1, 5, 5, 3.14159 ) AS rast ) AS rast FULL JOIN ( SELECT generate_series(1, 10) AS id ) AS id ON 1 = 1 ) SELECT ST_CountAgg(rast, 1, TRUE) FROM foo; st_countagg ------------- 20 (1 row)
ST_Histogram — Returns a set of record summarizing a raster or raster coverage data distribution separate bin ranges. Number of bins are autocomputed if not specified.
SETOF record ST_Histogram(
raster rast, integer nband=1, boolean exclude_nodata_value=true, integer bins=autocomputed, double precision[] width=NULL, boolean right=false)
;
SETOF record ST_Histogram(
raster rast, integer nband, integer bins, double precision[] width=NULL, boolean right=false)
;
SETOF record ST_Histogram(
raster rast, integer nband, boolean exclude_nodata_value, integer bins, boolean right)
;
SETOF record ST_Histogram(
raster rast, integer nband, integer bins, boolean right)
;
SETOF record ST_Histogram(
text rastertable, text rastercolumn, integer nband, integer bins, boolean right)
;
SETOF record ST_Histogram(
text rastertable, text rastercolumn, integer nband, boolean exclude_nodata_value, integer bins, boolean right)
;
SETOF record ST_Histogram(
text rastertable, text rastercolumn, integer nband=1, boolean exclude_nodata_value=true, integer bins=autocomputed, double precision[] width=NULL, boolean right=false)
;
SETOF record ST_Histogram(
text rastertable, text rastercolumn, integer nband=1, integer bins, double precision[] width=NULL, boolean right=false)
;
Returns set of records consisting of min, max, count, percent for a given raster band for each bin. If no band is specified nband
defaults to 1.
By default only considers pixel values not equal to the |
width
double precision[]width: an array indicating the width of each category/bin. If the number of bins is greater than the number of widths, the widths are repeated.
Example: 9 bins, widths are [a, b, c] will have the output be [a, b, c, a, b, c, a, b, c]
bins
integerNumber of breakouts -- this is the number of records you'll get back from the function if specified. If not specified then the number of breakouts is autocomputed.
right
booleancompute the histogram from the right rather than from the left (default). This changes the criteria for evaluating a value x from [a, b) to (a, b]
Disponibilità: 2.0
SELECT band, (stats).* FROM (SELECT rid, band, ST_Histogram(rast, band) As stats FROM dummy_rast CROSS JOIN generate_series(1,3) As band WHERE rid=2) As foo; band | min | max | count | percent ------+-------+-------+-------+--------- 1 | 249 | 250 | 2 | 0.08 1 | 250 | 251 | 2 | 0.08 1 | 251 | 252 | 1 | 0.04 1 | 252 | 253 | 2 | 0.08 1 | 253 | 254 | 18 | 0.72 2 | 78 | 113.2 | 11 | 0.44 2 | 113.2 | 148.4 | 4 | 0.16 2 | 148.4 | 183.6 | 4 | 0.16 2 | 183.6 | 218.8 | 1 | 0.04 2 | 218.8 | 254 | 5 | 0.2 3 | 62 | 100.4 | 11 | 0.44 3 | 100.4 | 138.8 | 5 | 0.2 3 | 138.8 | 177.2 | 4 | 0.16 3 | 177.2 | 215.6 | 1 | 0.04 3 | 215.6 | 254 | 4 | 0.16
SELECT (stats).* FROM (SELECT rid, ST_Histogram(rast, 2,6) As stats FROM dummy_rast WHERE rid=2) As foo; min | max | count | percent ------------+------------+-------+--------- 78 | 107.333333 | 9 | 0.36 107.333333 | 136.666667 | 6 | 0.24 136.666667 | 166 | 0 | 0 166 | 195.333333 | 4 | 0.16 195.333333 | 224.666667 | 1 | 0.04 224.666667 | 254 | 5 | 0.2 (6 rows) -- Same as previous but we explicitly control the pixel value range of each bin. SELECT (stats).* FROM (SELECT rid, ST_Histogram(rast, 2,6,ARRAY[0.5,1,4,100,5]) As stats FROM dummy_rast WHERE rid=2) As foo; min | max | count | percent -------+-------+-------+---------- 78 | 78.5 | 1 | 0.08 78.5 | 79.5 | 1 | 0.04 79.5 | 83.5 | 0 | 0 83.5 | 183.5 | 17 | 0.0068 183.5 | 188.5 | 0 | 0 188.5 | 254 | 6 | 0.003664 (6 rows)
ST_Quantile — Compute quantiles for a raster or raster table coverage in the context of the sample or population. Thus, a value could be examined to be at the raster's 25%, 50%, 75% percentile.
SETOF record ST_Quantile(
raster rast, integer nband=1, boolean exclude_nodata_value=true, double precision[] quantiles=NULL)
;
SETOF record ST_Quantile(
raster rast, double precision[] quantiles)
;
SETOF record ST_Quantile(
raster rast, integer nband, double precision[] quantiles)
;
double precision ST_Quantile(
raster rast, double precision quantile)
;
double precision ST_Quantile(
raster rast, boolean exclude_nodata_value, double precision quantile=NULL)
;
double precision ST_Quantile(
raster rast, integer nband, double precision quantile)
;
double precision ST_Quantile(
raster rast, integer nband, boolean exclude_nodata_value, double precision quantile)
;
double precision ST_Quantile(
raster rast, integer nband, double precision quantile)
;
SETOF record ST_Quantile(
text rastertable, text rastercolumn, integer nband=1, boolean exclude_nodata_value=true, double precision[] quantiles=NULL)
;
SETOF record ST_Quantile(
text rastertable, text rastercolumn, integer nband, double precision[] quantiles)
;
Compute quantiles for a raster or raster table coverage in the context of the sample or population. Thus, a value could be examined to be at the raster's 25%, 50%, 75% percentile.
If |
Disponibilità: 2.0
UPDATE dummy_rast SET rast = ST_SetBandNoDataValue(rast,249) WHERE rid=2; --Example will consider only pixels of band 1 that are not 249 and in named quantiles -- SELECT (pvq).* FROM (SELECT ST_Quantile(rast, ARRAY[0.25,0.75]) As pvq FROM dummy_rast WHERE rid=2) As foo ORDER BY (pvq).quantile; quantile | value ----------+------- 0.25 | 253 0.75 | 254 SELECT ST_Quantile(rast, 0.75) As value FROM dummy_rast WHERE rid=2; value ------ 254
--real live example. Quantile of all pixels in band 2 intersecting a geometry SELECT rid, (ST_Quantile(rast,2)).* As pvc FROM o_4_boston WHERE ST_Intersects(rast, ST_GeomFromText('POLYGON((224486 892151,224486 892200,224706 892200,224706 892151,224486 892151))',26986) ) ORDER BY value, quantile,rid ; rid | quantile | value -----+----------+------- 1 | 0 | 0 2 | 0 | 0 14 | 0 | 1 15 | 0 | 2 14 | 0.25 | 37 1 | 0.25 | 42 15 | 0.25 | 47 2 | 0.25 | 50 14 | 0.5 | 56 1 | 0.5 | 64 15 | 0.5 | 66 2 | 0.5 | 77 14 | 0.75 | 81 15 | 0.75 | 87 1 | 0.75 | 94 2 | 0.75 | 106 14 | 1 | 199 1 | 1 | 244 2 | 1 | 255 15 | 1 | 255
ST_SummaryStats — Returns summarystats consisting of count, sum, mean, stddev, min, max for a given raster band of a raster or raster coverage. Band 1 is assumed is no band is specified.
summarystats ST_SummaryStats(
raster rast, boolean exclude_nodata_value)
;
summarystats ST_SummaryStats(
raster rast, integer nband, boolean exclude_nodata_value)
;
summarystats ST_SummaryStats(
text rastertable, text rastercolumn, boolean exclude_nodata_value)
;
summarystats ST_SummaryStats(
text rastertable, text rastercolumn, integer nband=1, boolean exclude_nodata_value=true)
;
Returns summarystats consisting of count, sum, mean, stddev, min, max for a given raster band of a raster or raster coverage. If no band is specified nband
defaults to 1.
By default only considers pixel values not equal to the |
By default will sample all pixels. To get faster response, set |
Disponibilità: 2.0
The ST_SummaryStats(rastertable, rastercolumn, ...) variants are deprecated as of 2.2.0. Use ST_SummaryStatsAgg instead. |
SELECT rid, band, (stats).* FROM (SELECT rid, band, ST_SummaryStats(rast, band) As stats FROM dummy_rast CROSS JOIN generate_series(1,3) As band WHERE rid=2) As foo; rid | band | count | sum | mean | stddev | min | max -----+------+-------+------+------------+-----------+-----+----- 2 | 1 | 23 | 5821 | 253.086957 | 1.248061 | 250 | 254 2 | 2 | 25 | 3682 | 147.28 | 59.862188 | 78 | 254 2 | 3 | 25 | 3290 | 131.6 | 61.647384 | 62 | 254
This example took 574ms on PostGIS windows 64-bit with all of Boston Buildings and aerial Tiles (tiles each 150x150 pixels ~ 134,000 tiles), ~102,000 building records
WITH -- our features of interest feat AS (SELECT gid As building_id, geom_26986 As geom FROM buildings AS b WHERE gid IN(100, 103,150) ), -- clip band 2 of raster tiles to boundaries of builds -- then get stats for these clipped regions b_stats AS (SELECT building_id, (stats).* FROM (SELECT building_id, ST_SummaryStats(ST_Clip(rast,2,geom)) As stats FROM aerials.boston INNER JOIN feat ON ST_Intersects(feat.geom,rast) ) As foo ) -- finally summarize stats SELECT building_id, SUM(count) As num_pixels , MIN(min) As min_pval , MAX(max) As max_pval , SUM(mean*count)/SUM(count) As avg_pval FROM b_stats WHERE count > 0 GROUP BY building_id ORDER BY building_id; building_id | num_pixels | min_pval | max_pval | avg_pval -------------+------------+----------+----------+------------------ 100 | 1090 | 1 | 255 | 61.0697247706422 103 | 655 | 7 | 182 | 70.5038167938931 150 | 895 | 2 | 252 | 185.642458100559
-- stats for each band -- SELECT band, (stats).* FROM (SELECT band, ST_SummaryStats('o_4_boston','rast', band) As stats FROM generate_series(1,3) As band) As foo; band | count | sum | mean | stddev | min | max ------+---------+--------+------------------+------------------+-----+----- 1 | 8450000 | 725799 | 82.7064349112426 | 45.6800222638537 | 0 | 255 2 | 8450000 | 700487 | 81.4197705325444 | 44.2161184161765 | 0 | 255 3 | 8450000 | 575943 | 74.682739408284 | 44.2143885481407 | 0 | 255 -- For a table -- will get better speed if set sampling to less than 100% -- Here we set to 25% and get a much faster answer SELECT band, (stats).* FROM (SELECT band, ST_SummaryStats('o_4_boston','rast', band,true,0.25) As stats FROM generate_series(1,3) As band) As foo; band | count | sum | mean | stddev | min | max ------+---------+--------+------------------+------------------+-----+----- 1 | 2112500 | 180686 | 82.6890480473373 | 45.6961043857248 | 0 | 255 2 | 2112500 | 174571 | 81.448503668639 | 44.2252623171821 | 0 | 255 3 | 2112500 | 144364 | 74.6765884023669 | 44.2014869384578 | 0 | 255
ST_SummaryStatsAgg — Aggregate. Returns summarystats consisting of count, sum, mean, stddev, min, max for a given raster band of a set of raster. Band 1 is assumed is no band is specified.
summarystats ST_SummaryStatsAgg(
raster rast, integer nband, boolean exclude_nodata_value, double precision sample_percent)
;
summarystats ST_SummaryStatsAgg(
raster rast, boolean exclude_nodata_value, double precision sample_percent)
;
summarystats ST_SummaryStatsAgg(
raster rast, integer nband, boolean exclude_nodata_value)
;
Returns summarystats consisting of count, sum, mean, stddev, min, max for a given raster band of a raster or raster coverage. If no band is specified nband
defaults to 1.
By default only considers pixel values not equal to the |
By default will sample all pixels. To get faster response, set |
Availability: 2.2.0
WITH foo AS ( SELECT rast.rast FROM ( SELECT ST_SetValue( ST_SetValue( ST_SetValue( ST_AddBand( ST_MakeEmptyRaster(10, 10, 10, 10, 2, 2, 0, 0,0) , 1, '64BF', 0, 0 ) , 1, 1, 1, -10 ) , 1, 5, 4, 0 ) , 1, 5, 5, 3.14159 ) AS rast ) AS rast FULL JOIN ( SELECT generate_series(1, 10) AS id ) AS id ON 1 = 1 ) SELECT (stats).count, round((stats).sum::numeric, 3), round((stats).mean::numeric, 3), round((stats).stddev::numeric, 3), round((stats).min::numeric, 3), round((stats).max::numeric, 3) FROM ( SELECT ST_SummaryStatsAgg(rast, 1, TRUE, 1) AS stats FROM foo ) bar; count | round | round | round | round | round -------+---------+--------+-------+---------+------- 20 | -68.584 | -3.429 | 6.571 | -10.000 | 3.142 (1 row)
ST_ValueCount — Returns a set of records containing a pixel band value and count of the number of pixels in a given band of a raster (or a raster coverage) that have a given set of values. If no band is specified defaults to band 1. By default nodata value pixels are not counted. and all other values in the pixel are output and pixel band values are rounded to the nearest integer.
SETOF record ST_ValueCount(
raster rast, integer nband=1, boolean exclude_nodata_value=true, double precision[] searchvalues=NULL, double precision roundto=0, double precision OUT value, integer OUT count)
;
SETOF record ST_ValueCount(
raster rast, integer nband, double precision[] searchvalues, double precision roundto=0, double precision OUT value, integer OUT count)
;
SETOF record ST_ValueCount(
raster rast, double precision[] searchvalues, double precision roundto=0, double precision OUT value, integer OUT count)
;
bigint ST_ValueCount(
raster rast, double precision searchvalue, double precision roundto=0)
;
bigint ST_ValueCount(
raster rast, integer nband, boolean exclude_nodata_value, double precision searchvalue, double precision roundto=0)
;
bigint ST_ValueCount(
raster rast, integer nband, double precision searchvalue, double precision roundto=0)
;
SETOF record ST_ValueCount(
text rastertable, text rastercolumn, integer nband=1, boolean exclude_nodata_value=true, double precision[] searchvalues=NULL, double precision roundto=0, double precision OUT value, integer OUT count)
;
SETOF record ST_ValueCount(
text rastertable, text rastercolumn, double precision[] searchvalues, double precision roundto=0, double precision OUT value, integer OUT count)
;
SETOF record ST_ValueCount(
text rastertable, text rastercolumn, integer nband, double precision[] searchvalues, double precision roundto=0, double precision OUT value, integer OUT count)
;
bigintST_ValueCount(
text rastertable, text rastercolumn, integer nband, boolean exclude_nodata_value, double precision searchvalue, double precision roundto=0)
;
bigint ST_ValueCount(
text rastertable, text rastercolumn, double precision searchvalue, double precision roundto=0)
;
bigint ST_ValueCount(
text rastertable, text rastercolumn, integer nband, double precision searchvalue, double precision roundto=0)
;
Returns a set of records with columns value
count
which contain the pixel band value and count of pixels in the raster tile or raster coverage of selected band.
If no band is specified nband
defaults to 1. If no searchvalues
are specified, will return all pixel values found in the raster or raster coverage. If one searchvalue is given, will return an integer instead of records denoting the count of pixels having that pixel band value
If |
Disponibilità: 2.0
UPDATE dummy_rast SET rast = ST_SetBandNoDataValue(rast,249) WHERE rid=2; --Example will count only pixels of band 1 that are not 249. -- SELECT (pvc).* FROM (SELECT ST_ValueCount(rast) As pvc FROM dummy_rast WHERE rid=2) As foo ORDER BY (pvc).value; value | count -------+------- 250 | 2 251 | 1 252 | 2 253 | 6 254 | 12 -- Example will coount all pixels of band 1 including 249 -- SELECT (pvc).* FROM (SELECT ST_ValueCount(rast,1,false) As pvc FROM dummy_rast WHERE rid=2) As foo ORDER BY (pvc).value; value | count -------+------- 249 | 2 250 | 2 251 | 1 252 | 2 253 | 6 254 | 12 -- Example will count only non-nodata value pixels of band 2 SELECT (pvc).* FROM (SELECT ST_ValueCount(rast,2) As pvc FROM dummy_rast WHERE rid=2) As foo ORDER BY (pvc).value; value | count -------+------- 78 | 1 79 | 1 88 | 1 89 | 1 96 | 1 97 | 1 98 | 1 99 | 2 112 | 2 :
--real live example. Count all the pixels in an aerial raster tile band 2 intersecting a geometry -- and return only the pixel band values that have a count > 500 SELECT (pvc).value, SUM((pvc).count) As total FROM (SELECT ST_ValueCount(rast,2) As pvc FROM o_4_boston WHERE ST_Intersects(rast, ST_GeomFromText('POLYGON((224486 892151,224486 892200,224706 892200,224706 892151,224486 892151))',26986) ) ) As foo GROUP BY (pvc).value HAVING SUM((pvc).count) > 500 ORDER BY (pvc).value; value | total -------+----- 51 | 502 54 | 521
-- Just return count of pixels in each raster tile that have value of 100 of tiles that intersect a specific geometry -- SELECT rid, ST_ValueCount(rast,2,100) As count FROM o_4_boston WHERE ST_Intersects(rast, ST_GeomFromText('POLYGON((224486 892151,224486 892200,224706 892200,224706 892151,224486 892151))',26986) ) ; rid | count -----+------- 1 | 56 2 | 95 14 | 37 15 | 64
ST_AsBinary — Return the Well-Known Binary (WKB) representation of the raster without SRID meta data.
bytea ST_AsBinary(
raster rast, boolean outasin=FALSE)
;
Returns the Binary representation of the raster. If outasin
is TRUE, out-db bands are treated as in-db.
This is useful in binary cursors to pull data out of the database without converting it to a string representation.
By default, WKB output contains the external file path for out-db bands. If the client does not have access to the raster file underlying an out-db band, set |
Enhanced: 2.1.0 Addition of outasin
SELECT ST_AsBinary(rast) As rastbin FROM dummy_rast WHERE rid=1; rastbin --------------------------------------------------------------------------------- \001\000\000\000\000\000\000\000\000\000\000\000@\000\000\000\000\000\000\010@\ 000\000\000\000\000\000\340?\000\000\000\000\000\000\340?\000\000\000\000\000\00 0\000\000\000\000\000\000\000\000\000\000\012\000\000\000\012\000\024\000
ST_AsGDALRaster — Return the raster tile in the designated GDAL Raster format. Raster formats are one of those supported by your compiled library. Use ST_GDALRasters() to get a list of formats supported by your library.
bytea ST_AsGDALRaster(
raster rast, text format, text[] options=NULL, integer srid=sameassource)
;
Returns the raster tile in the designated format. Arguments are itemized below:
format
format to output. This is dependent on the drivers compiled in your libgdal library. Generally available are 'JPEG', 'GTIff', 'PNG'. Use ST_GDALDrivers to get a list of formats supported by your library.
options
text array of GDAL options. Valid options are dependent on the format. Refer to GDAL Raster format options for more details.
srs
The proj4text or srtext (from spatial_ref_sys) to embed in the image
Availability: 2.0.0 - requires GDAL >= 1.6.0.
SELECT ST_AsGDALRaster(rast, 'JPEG') As rastjpg FROM dummy_rast WHERE rid=1; SELECT ST_AsGDALRaster(rast, 'JPEG', ARRAY['QUALITY=50']) As rastjpg FROM dummy_rast WHERE rid=2;
ST_AsJPEG — Return the raster tile selected bands as a single Joint Photographic Exports Group (JPEG) image (byte array). If no band is specified and 1 or more than 3 bands, then only the first band is used. If only 3 bands then all 3 bands are used and mapped to RGB.
bytea ST_AsJPEG(
raster rast, text[] options=NULL)
;
bytea ST_AsJPEG(
raster rast, integer nband, integer quality)
;
bytea ST_AsJPEG(
raster rast, integer nband, text[] options=NULL)
;
bytea ST_AsJPEG(
raster rast, integer[] nbands, text[] options=NULL)
;
bytea ST_AsJPEG(
raster rast, integer[] nbands, integer quality)
;
Returns the selected bands of the raster as a single Joint Photographic Exports Group Image (JPEG). Use ST_AsGDALRaster if you need to export as less common raster types. If no band is specified and 1 or more than 3 bands, then only the first band is used. If 3 bands then all 3 bands are used. There are many variants of the function with many options. These are itemized below:
nband
is for single band exports.
nbands
is an array of bands to export (note that max is 3 for JPEG) and the order of the bands is RGB. e.g ARRAY[3,2,1] means map band 3 to Red, band 2 to green and band 1 to blue
quality
number from 0 to 100. The higher the number the crisper the image.
options
text Array of GDAL options as defined for JPEG (look at create_options for JPEG ST_GDALDrivers). For JPEG valid ones are PROGRESSIVE
ON or OFF and QUALITY
a range from 0 to 100 and default to 75. Refer to GDAL Raster format options for more details.
Availability: 2.0.0 - requires GDAL >= 1.6.0.
-- output first 3 bands 75% quality SELECT ST_AsJPEG(rast) As rastjpg FROM dummy_rast WHERE rid=2; -- output only first band as 90% quality SELECT ST_AsJPEG(rast,1,90) As rastjpg FROM dummy_rast WHERE rid=2; -- output first 3 bands (but make band 2 Red, band 1 green, and band 3 blue, progressive and 90% quality SELECT ST_AsJPEG(rast,ARRAY[2,1,3],ARRAY['QUALITY=90','PROGRESSIVE=ON']) As rastjpg FROM dummy_rast WHERE rid=2;
ST_AsPNG — Return the raster tile selected bands as a single portable network graphics (PNG) image (byte array). If 1, 3, or 4 bands in raster and no bands are specified, then all bands are used. If more 2 or more than 4 bands and no bands specified, then only band 1 is used. Bands are mapped to RGB or RGBA space.
bytea ST_AsPNG(
raster rast, text[] options=NULL)
;
bytea ST_AsPNG(
raster rast, integer nband, integer compression)
;
bytea ST_AsPNG(
raster rast, integer nband, text[] options=NULL)
;
bytea ST_AsPNG(
raster rast, integer[] nbands, integer compression)
;
bytea ST_AsPNG(
raster rast, integer[] nbands, text[] options=NULL)
;
Returns the selected bands of the raster as a single Portable Network Graphics Image (PNG). Use ST_AsGDALRaster if you need to export as less common raster types. If no band is specified, then the first 3 bands are exported. There are many variants of the function with many options. If no srid
is specified then then srid of the raster is used. These are itemized below:
nband
is for single band exports.
nbands
is an array of bands to export (note that max is 3 for PNG) and the order of the bands is RGB. e.g ARRAY[3,2,1] means map band 3 to Red, band 2 to green and band 1 to blue
compression
number from 1 to 9. The higher the number the greater the compression.
options
text Array of GDAL options as defined for PNG (look at create_options for PNG of ST_GDALDrivers). For PNG valid one is only ZLEVEL (amount of time to spend on compression -- default 6) e.g. ARRAY['ZLEVEL=9']. WORLDFILE is not allowed since the function would have to output two outputs. Refer to GDAL Raster format options for more details.
Availability: 2.0.0 - requires GDAL >= 1.6.0.
ST_AsTIFF — Return the raster selected bands as a single TIFF image (byte array). If no band is specified, then will try to use all bands.
bytea ST_AsTIFF(
raster rast, text[] options='', integer srid=sameassource)
;
bytea ST_AsTIFF(
raster rast, text compression='', integer srid=sameassource)
;
bytea ST_AsTIFF(
raster rast, integer[] nbands, text compression='', integer srid=sameassource)
;
bytea ST_AsTIFF(
raster rast, integer[] nbands, text[] options, integer srid=sameassource)
;
Returns the selected bands of the raster as a single Tagged Image File Format (TIFF). If no band is specified, will try to use all bands. This is a wrapper around ST_AsGDALRaster. Use ST_AsGDALRaster if you need to export as less common raster types. There are many variants of the function with many options. If no spatial reference SRS text is present, the spatial reference of the raster is used. These are itemized below:
nbands
is an array of bands to export (note that max is 3 for PNG) and the order of the bands is RGB. e.g ARRAY[3,2,1] means map band 3 to Red, band 2 to green and band 1 to blue
compression
Compression expression -- JPEG90 (or some other percent), LZW, JPEG, DEFLATE9.
options
text Array of GDAL create options as defined for GTiff (look at create_options for GTiff of ST_GDALDrivers). or refer to GDAL Raster format options for more details.
srid
srid of spatial_ref_sys of the raster. This is used to populate the georeference information
Availability: 2.0.0 - requires GDAL >= 1.6.0.
crop
is not specified or TRUE, the output raster is cropped.ST_Clip — Returns the raster clipped by the input geometry. If band number not is specified, all bands are processed. If crop
is not specified or TRUE, the output raster is cropped.
raster ST_Clip(
raster rast, integer[] nband, geometry geom, double precision[] nodataval=NULL, boolean crop=TRUE)
;
raster ST_Clip(
raster rast, integer nband, geometry geom, double precision nodataval, boolean crop=TRUE)
;
raster ST_Clip(
raster rast, integer nband, geometry geom, boolean crop)
;
raster ST_Clip(
raster rast, geometry geom, double precision[] nodataval=NULL, boolean crop=TRUE)
;
raster ST_Clip(
raster rast, geometry geom, double precision nodataval, boolean crop=TRUE)
;
raster ST_Clip(
raster rast, geometry geom, boolean crop)
;
Returns a raster that is clipped by the input geometry geom
. If band index is not specified, all bands are processed.
Rasters resulting from ST_Clip must have a nodata value assigned for areas clipped, one for each band. If none are provided and the input raster do not have a nodata value defined, nodata values of the resulting raster are set to ST_MinPossibleValue(ST_BandPixelType(rast, band)). When the number of nodata value in the array is smaller than the number of band, the last one in the array is used for the remaining bands. If the number of nodata value is greater than the number of band, the extra nodata values are ignored. All variants accepting an array of nodata values also accept a single value which will be assigned to each band.
If crop
is not specified, true is assumed meaning the output raster is cropped to the intersection of the geom
and rast
extents. If crop
is set to false, the new raster gets the same extent as rast
.
Disponibilità: 2.0
Enhanced: 2.1.0 Rewritten in C
Examples here use Massachusetts aerial data available on MassGIS site MassGIS Aerial Orthos. Coordinates are in Massachusetts State Plane Meters.
-- Clip the first band of an aerial tile by a 20 meter buffer. SELECT ST_Clip(rast, 1, ST_Buffer(ST_Centroid(ST_Envelope(rast)),20) ) from aerials.boston WHERE rid = 4;
-- Demonstrate effect of crop on final dimensions of raster -- Note how final extent is clipped to that of the geometry -- if crop = true SELECT ST_XMax(ST_Envelope(ST_Clip(rast, 1, clipper, true))) As xmax_w_trim, ST_XMax(clipper) As xmax_clipper, ST_XMax(ST_Envelope(ST_Clip(rast, 1, clipper, false))) As xmax_wo_trim, ST_XMax(ST_Envelope(rast)) As xmax_rast_orig FROM (SELECT rast, ST_Buffer(ST_Centroid(ST_Envelope(rast)),6) As clipper FROM aerials.boston WHERE rid = 6) As foo; xmax_w_trim | xmax_clipper | xmax_wo_trim | xmax_rast_orig ------------------+------------------+------------------+------------------ 230657.436173996 | 230657.436173996 | 230666.436173996 | 230666.436173996
|
|
-- Same example as before, but we need to set crop to false to be able to use ST_AddBand -- because ST_AddBand requires all bands be the same Width and height SELECT ST_AddBand(ST_Clip(rast, 1, ST_Buffer(ST_Centroid(ST_Envelope(rast)),20),false ), ARRAY[ST_Band(rast,2),ST_Band(rast,3)] ) from aerials.boston WHERE rid = 6;
|
|
ST_ColorMap — Creates a new raster of up to four 8BUI bands (grayscale, RGB, RGBA) from the source raster and a specified band. Band 1 is assumed if not specified.
raster ST_ColorMap(
raster rast, integer nband=1, text colormap=grayscale, text method=INTERPOLATE)
;
raster ST_ColorMap(
raster rast, text colormap, text method=INTERPOLATE)
;
Apply a colormap
to the band at nband
of rast
resulting a new raster comprised of up to four 8BUI bands. The number of 8BUI bands in the new raster is determined by the number of color components defined in colormap
.
If nband
is not specified, then band 1 is assumed.
colormap
can be a keyword of a pre-defined colormap or a set of lines defining the value and the color components.
Valid pre-defined colormap
keyword:
grayscale
or greyscale
for a one 8BUI band raster of shades of gray.
pseudocolor
for a four 8BUI (RGBA) band raster with colors going from blue to green to red.
fire
for a four 8BUI (RGBA) band raster with colors going from black to red to pale yellow.
bluered
for a four 8BUI (RGBA) band raster with colors going from blue to pale white to red.
Users can pass a set of entries (one per line) to colormap
to specify custom colormaps. Each entry generally consists of five values: the pixel value and corresponding Red, Green, Blue, Alpha components (color components between 0 and 255). Percent values can be used instead of pixel values where 0% and 100% are the minimum and maximum values found in the raster band. Values can be separated with commas (','), tabs, colons (':') and/or spaces. The pixel value can be set to nv, null or nodata for the NODATA value. An example is provided below.
5 0 0 0 255 4 100:50 55 255 1 150,100 150 255 0% 255 255 255 255 nv 0 0 0 0
The syntax of colormap
is similar to that of the color-relief mode of GDAL gdaldem.
Valid keywords for method
:
INTERPOLATE
to use linear interpolation to smoothly blend the colors between the given pixel values
EXACT
to strictly match only those pixels values found in the colormap. Pixels whose value does not match a colormap entry will be set to 0 0 0 0 (RGBA)
NEAREST
to use the colormap entry whose value is closest to the pixel value
A great reference for colormaps is ColorBrewer. |
The resulting bands of new raster will have no NODATA value set. Use ST_SetBandNoDataValue to set a NODATA value if one is needed. |
Availability: 2.1.0
This is a junk table to play with
-- setup test raster table -- DROP TABLE IF EXISTS funky_shapes; CREATE TABLE funky_shapes(rast raster); INSERT INTO funky_shapes(rast) WITH ref AS ( SELECT ST_MakeEmptyRaster( 200, 200, 0, 200, 1, -1, 0, 0) AS rast ) SELECT ST_Union(rast) FROM ( SELECT ST_AsRaster( ST_Rotate( ST_Buffer( ST_GeomFromText('LINESTRING(0 2,50 50,150 150,125 50)'), i*2 ), pi() * i * 0.125, ST_Point(50,50) ), ref.rast, '8BUI'::text, i * 5 ) AS rast FROM ref CROSS JOIN generate_series(1, 10, 3) AS i ) AS shapes;
SELECT ST_NumBands(rast) As n_orig, ST_NumBands(ST_ColorMap(rast,1, 'greyscale')) As ngrey, ST_NumBands(ST_ColorMap(rast,1, 'pseudocolor')) As npseudo, ST_NumBands(ST_ColorMap(rast,1, 'fire')) As nfire, ST_NumBands(ST_ColorMap(rast,1, 'bluered')) As nbluered, ST_NumBands(ST_ColorMap(rast,1, ' 100% 255 0 0 80% 160 0 0 50% 130 0 0 30% 30 0 0 20% 60 0 0 0% 0 0 0 nv 255 255 255 ')) As nred FROM funky_shapes;
n_orig | ngrey | npseudo | nfire | nbluered | nred --------+-------+---------+-------+----------+------ 1 | 1 | 4 | 4 | 4 | 3
SELECT ST_AsPNG(rast) As orig_png, ST_AsPNG(ST_ColorMap(rast,1,'greyscale')) As grey_png, ST_AsPNG(ST_ColorMap(rast,1, 'pseudocolor')) As pseudo_png, ST_AsPNG(ST_ColorMap(rast,1, 'nfire')) As fire_png, ST_AsPNG(ST_ColorMap(rast,1, 'bluered')) As bluered_png, ST_AsPNG(ST_ColorMap(rast,1, ' 100% 255 0 0 80% 160 0 0 50% 130 0 0 30% 30 0 0 20% 60 0 0 0% 0 0 0 nv 255 255 255 ')) As red_png FROM funky_shapes;
|
|
|
|
|
|
ST_AsPNG, ST_AsRaster ST_MapAlgebra, ST_NumBands, ST_Reclass, ST_SetBandNoDataValue, ST_Union
ST_Intersection — Returns a raster or a set of geometry-pixelvalue pairs representing the shared portion of two rasters or the geometrical intersection of a vectorization of the raster and a geometry.
setof geomval ST_Intersection(
geometry geom, raster rast, integer band_num=1)
;
setof geomval ST_Intersection(
raster rast, geometry geom)
;
setof geomval ST_Intersection(
raster rast, integer band_num, geometry geom)
;
raster ST_Intersection(
raster rast1, raster rast2, double precision[] nodataval)
;
raster ST_Intersection(
raster rast1, raster rast2, text returnband='BOTH', double precision[] nodataval=NULL)
;
raster ST_Intersection(
raster rast1, integer band_num1, raster rast2, integer band_num2, double precision[] nodataval)
;
raster ST_Intersection(
raster rast1, integer band_num1, raster rast2, integer band_num2, text returnband='BOTH', double precision[] nodataval=NULL)
;
Returns a raster or a set of geometry-pixelvalue pairs representing the shared portion of two rasters or the geometrical intersection of a vectorization of the raster and a geometry.
The first three variants, returning a setof geomval, works in vector space. The raster is first vectorized (using ST_DumpAsPolygon) into a set of geomval rows and those rows are then intersected with the geometry using the ST_Intersection(geometry, geometry) PostGIS function. Geometries intersecting only with a nodata value area of a raster returns an empty geometry. They are normally excluded from the results by the proper usage of ST_Intersect in the WHERE clause.
You can access the geometry and the value parts of the resulting set of geomval by surrounding them with parenthesis and adding '.geom' or '.val' at the end of the expression. e.g. (ST_Intersection(rast, geom)).geom
The other variants, returning a raster, works in raster space. They are using the two rasters version of ST_MapAlgebraExpr to perform the intersection.
The extent of the resulting raster corresponds to the geometrical intersection of the two raster extents. The resulting raster includes 'BAND1', 'BAND2' or 'BOTH' bands, following what is passed as the returnband
parameter. Nodata value areas present in any band results in nodata value areas in every bands of the result. In other words, any pixel intersecting with a nodata value pixel becomes a nodata value pixel in the result.
Rasters resulting from ST_Intersection must have a nodata value assigned for areas not intersecting. You can define or replace the nodata value for any resulting band by providing a nodataval[]
array of one or two nodata values depending if you request 'BAND1', 'BAND2' or 'BOTH' bands. The first value in the array replace the nodata value in the first band and the second value replace the nodata value in the second band. If one input band do not have a nodata value defined and none are provided as an array, one is chosen using the ST_MinPossibleValue function. All variant accepting an array of nodata value can also accept a single value which will be assigned to each requested band.
In all variants, if no band number is specified band 1 is assumed. If you need an intersection between a raster and geometry that returns a raster, refer to ST_Clip.
To get more control on the resulting extent or on what to return when encountering a nodata value, use the two rasters version of ST_MapAlgebraExpr. |
To compute the intersection of a raster band with a geometry in raster space, use ST_Clip. ST_Clip works on multiple bands rasters and does not return a band corresponding to the rasterized geometry. |
ST_Intersection should be used in conjunction with ST_Intersects and an index on the raster column and/or the geometry column. |
Enhanced: 2.0.0 - Intersection in the raster space was introduced. In earlier pre-2.0.0 versions, only intersection performed in vector space were supported.
SELECT foo.rid, foo.gid, ST_AsText((foo.geomval).geom) As geomwkt, (foo.geomval).val FROM ( SELECT A.rid, g.gid, ST_Intersection(A.rast, g.geom) As geomval FROM dummy_rast AS A CROSS JOIN ( VALUES (1, ST_Point(3427928, 5793243.85) ), (2, ST_GeomFromText('LINESTRING(3427927.85 5793243.75,3427927.8 5793243.75,3427927.8 5793243.8)')), (3, ST_GeomFromText('LINESTRING(1 2, 3 4)')) ) As g(gid,geom) WHERE A.rid = 2 ) As foo; rid | gid | geomwkt | val -----+-----+--------------------------------------------------------------------------------------------- 2 | 1 | POINT(3427928 5793243.85) | 249 2 | 1 | POINT(3427928 5793243.85) | 253 2 | 2 | POINT(3427927.85 5793243.75) | 254 2 | 2 | POINT(3427927.8 5793243.8) | 251 2 | 2 | POINT(3427927.8 5793243.8) | 253 2 | 2 | LINESTRING(3427927.8 5793243.75,3427927.8 5793243.8) | 252 2 | 2 | MULTILINESTRING((3427927.8 5793243.8,3427927.8 5793243.75),...) | 250 2 | 3 | GEOMETRYCOLLECTION EMPTY
ST_MapAlgebra — Callback function version - Returns a one-band raster given one or more input rasters, band indexes and one user-specified callback function.
raster ST_MapAlgebra(
rastbandarg[] rastbandargset, regprocedure callbackfunc, text pixeltype=NULL, text extenttype=INTERSECTION, raster customextent=NULL, integer distancex=0, integer distancey=0, text[] VARIADIC userargs=NULL)
;
raster ST_MapAlgebra(
raster rast, integer[] nband, regprocedure callbackfunc, text pixeltype=NULL, text extenttype=FIRST, raster customextent=NULL, integer distancex=0, integer distancey=0, text[] VARIADIC userargs=NULL)
;
raster ST_MapAlgebra(
raster rast, integer nband, regprocedure callbackfunc, text pixeltype=NULL, text extenttype=FIRST, raster customextent=NULL, integer distancex=0, integer distancey=0, text[] VARIADIC userargs=NULL)
;
raster ST_MapAlgebra(
raster rast1, integer nband1, raster rast2, integer nband2, regprocedure callbackfunc, text pixeltype=NULL, text extenttype=INTERSECTION, raster customextent=NULL, integer distancex=0, integer distancey=0, text[] VARIADIC userargs=NULL)
;
Returns a one-band raster given one or more input rasters, band indexes and one user-specified callback function.
Rasters on which the map algebra process is evaluated.
rastbandargset
allows the use of a map algebra operation on many rasters and/or many bands. See example Variant 1.
Band numbers of the raster to be evaluated. nband can be an integer or integer[] denoting the bands. nband1 is band on rast1 and nband2 is band on rast2 for hte 2 raster/2band case.
The callbackfunc
parameter must be the name and signature of an SQL or PL/pgSQL function, cast to a regprocedure. An example PL/pgSQL function example is:
CREATE OR REPLACE FUNCTION sample_callbackfunc(value double precision[][][], position integer[][], VARIADIC userargs text[]) RETURNS double precision AS $$ BEGIN RETURN 0; END; $$ LANGUAGE 'plpgsql' IMMUTABLE;
The callbackfunc
must have three arguments: a 3-dimension double precision array, a 2-dimension integer array and a variadic 1-dimension text array. The first argument value
is the set of values (as double precision) from all input rasters. The three dimensions (where indexes are 1-based) are: raster #, row y, column x. The second argument position
is the set of pixel positions from the output raster and input rasters. The outer dimension (where indexes are 0-based) is the raster #. The position at outer dimension index 0 is the output raster's pixel position. For each outer dimension, there are two elements in the inner dimension for X and Y. The third argument userargs
is for passing through any user-specified arguments.
Passing a regprocedure argument to a SQL function requires the full function signature to be passed, then cast to a regprocedure type. To pass the above example PL/pgSQL function as an argument, the SQL for the argument is:
'sample_callbackfunc(double precision[], integer[], text[])'::regprocedure
Note that the argument contains the name of the function, the types of the function arguments, quotes around the name and argument types, and a cast to a regprocedure.
An n-dimenional array (matrix) of numbers used to filter what cells get passed to map algebra call-back function. 0 means a neighbor cell value should be treated as no-data and 1 means value should be treated as data. If weight is set to true, then the values, are used as multipliers to multiple the pixel value of that value in the neighborhood position.
boolean (true/false) to denote if a mask value should be weighted (multiplied by original value) or not (only applies to proto that takes a mask).
If pixeltype
is passed in, the one band of the new raster will be of that pixeltype. If pixeltype is passed NULL or left out, the new raster band will have the same pixeltype as the specified band of the first raster (for extent types: INTERSECTION, UNION, FIRST, CUSTOM) or the specified band of the appropriate raster (for extent types: SECOND, LAST). If in doubt, always specify pixeltype
.
The resulting pixel type of the output raster must be one listed in ST_BandPixelType or left out or set to NULL.
Possible values are INTERSECTION (default), UNION, FIRST (default for one raster variants), SECOND, LAST, CUSTOM.
If extentype
is CUSTOM, a raster must be provided for customextent
. See example 4 of Variant 1.
The distance in pixels from the reference cell. So width of resulting matrix would be 2*distancex + 1
.If not specified only the reference cell is considered (neighborhood of 0).
The distance in pixels from reference cell in y direction. Height of resulting matrix would be 2*distancey + 1
.If not specified only the reference cell is considered (neighborhood of 0).
The third argument to the callbackfunc
is a variadic text array. All trailing text arguments are passed through to the specified callbackfunc
, and are contained in the userargs
argument.
For more information about the VARIADIC keyword, please refer to the PostgreSQL documentation and the "SQL Functions with Variable Numbers of Arguments" section of Query Language (SQL) Functions. |
The text[] argument to the |
Variant 1 accepts an array of rastbandarg
allowing the use of a map algebra operation on many rasters and/or many bands. See example Variant 1.
Variants 2 and 3 operate upon one or more bands of one raster. See example Variant 2 and 3.
Variant 4 operate upon two rasters with one band per raster. See example Variant 4.
Availability: 2.2.0: Ability to add a mask
Availability: 2.1.0
One raster, one band
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0) AS rast ) SELECT ST_MapAlgebra( ARRAY[ROW(rast, 1)]::rastbandarg[], 'sample_callbackfunc(double precision[], int[], text[])'::regprocedure ) AS rast FROM foo
One raster, several bands
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0), 2, '8BUI', 10, 0), 3, '32BUI', 100, 0) AS rast ) SELECT ST_MapAlgebra( ARRAY[ROW(rast, 3), ROW(rast, 1), ROW(rast, 3), ROW(rast, 2)]::rastbandarg[], 'sample_callbackfunc(double precision[], int[], text[])'::regprocedure ) AS rast FROM foo
Several rasters, several bands
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0), 2, '8BUI', 10, 0), 3, '32BUI', 100, 0) AS rast UNION ALL SELECT 2 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 1, 1, -1, 0, 0, 0), 1, '16BUI', 2, 0), 2, '8BUI', 20, 0), 3, '32BUI', 300, 0) AS rast ) SELECT ST_MapAlgebra( ARRAY[ROW(t1.rast, 3), ROW(t2.rast, 1), ROW(t2.rast, 3), ROW(t1.rast, 2)]::rastbandarg[], 'sample_callbackfunc(double precision[], int[], text[])'::regprocedure ) AS rast FROM foo t1 CROSS JOIN foo t2 WHERE t1.rid = 1 AND t2.rid = 2
Complete example of tiles of a coverage with neighborhood. This query only works with PostgreSQL 9.1 or higher.
WITH foo AS ( SELECT 0 AS rid, ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0) AS rast UNION ALL SELECT 1, ST_AddBand(ST_MakeEmptyRaster(2, 2, 2, 0, 1, -1, 0, 0, 0), 1, '16BUI', 2, 0) AS rast UNION ALL SELECT 2, ST_AddBand(ST_MakeEmptyRaster(2, 2, 4, 0, 1, -1, 0, 0, 0), 1, '16BUI', 3, 0) AS rast UNION ALL SELECT 3, ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, -2, 1, -1, 0, 0, 0), 1, '16BUI', 10, 0) AS rast UNION ALL SELECT 4, ST_AddBand(ST_MakeEmptyRaster(2, 2, 2, -2, 1, -1, 0, 0, 0), 1, '16BUI', 20, 0) AS rast UNION ALL SELECT 5, ST_AddBand(ST_MakeEmptyRaster(2, 2, 4, -2, 1, -1, 0, 0, 0), 1, '16BUI', 30, 0) AS rast UNION ALL SELECT 6, ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, -4, 1, -1, 0, 0, 0), 1, '16BUI', 100, 0) AS rast UNION ALL SELECT 7, ST_AddBand(ST_MakeEmptyRaster(2, 2, 2, -4, 1, -1, 0, 0, 0), 1, '16BUI', 200, 0) AS rast UNION ALL SELECT 8, ST_AddBand(ST_MakeEmptyRaster(2, 2, 4, -4, 1, -1, 0, 0, 0), 1, '16BUI', 300, 0) AS rast ) SELECT t1.rid, ST_MapAlgebra( ARRAY[ROW(ST_Union(t2.rast), 1)]::rastbandarg[], 'sample_callbackfunc(double precision[], int[], text[])'::regprocedure, '32BUI', 'CUSTOM', t1.rast, 1, 1 ) AS rast FROM foo t1 CROSS JOIN foo t2 WHERE t1.rid = 4 AND t2.rid BETWEEN 0 AND 8 AND ST_Intersects(t1.rast, t2.rast) GROUP BY t1.rid, t1.rast
Example like the prior one for tiles of a coverage with neighborhood but works with PostgreSQL 9.0.
WITH src AS ( SELECT 0 AS rid, ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0) AS rast UNION ALL SELECT 1, ST_AddBand(ST_MakeEmptyRaster(2, 2, 2, 0, 1, -1, 0, 0, 0), 1, '16BUI', 2, 0) AS rast UNION ALL SELECT 2, ST_AddBand(ST_MakeEmptyRaster(2, 2, 4, 0, 1, -1, 0, 0, 0), 1, '16BUI', 3, 0) AS rast UNION ALL SELECT 3, ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, -2, 1, -1, 0, 0, 0), 1, '16BUI', 10, 0) AS rast UNION ALL SELECT 4, ST_AddBand(ST_MakeEmptyRaster(2, 2, 2, -2, 1, -1, 0, 0, 0), 1, '16BUI', 20, 0) AS rast UNION ALL SELECT 5, ST_AddBand(ST_MakeEmptyRaster(2, 2, 4, -2, 1, -1, 0, 0, 0), 1, '16BUI', 30, 0) AS rast UNION ALL SELECT 6, ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, -4, 1, -1, 0, 0, 0), 1, '16BUI', 100, 0) AS rast UNION ALL SELECT 7, ST_AddBand(ST_MakeEmptyRaster(2, 2, 2, -4, 1, -1, 0, 0, 0), 1, '16BUI', 200, 0) AS rast UNION ALL SELECT 8, ST_AddBand(ST_MakeEmptyRaster(2, 2, 4, -4, 1, -1, 0, 0, 0), 1, '16BUI', 300, 0) AS rast ) WITH foo AS ( SELECT t1.rid, ST_Union(t2.rast) AS rast FROM src t1 JOIN src t2 ON ST_Intersects(t1.rast, t2.rast) AND t2.rid BETWEEN 0 AND 8 WHERE t1.rid = 4 GROUP BY t1.rid ), bar AS ( SELECT t1.rid, ST_MapAlgebra( ARRAY[ROW(t2.rast, 1)]::rastbandarg[], 'raster_nmapalgebra_test(double precision[], int[], text[])'::regprocedure, '32BUI', 'CUSTOM', t1.rast, 1, 1 ) AS rast FROM src t1 JOIN foo t2 ON t1.rid = t2.rid ) SELECT rid, (ST_Metadata(rast)), (ST_BandMetadata(rast, 1)), ST_Value(rast, 1, 1, 1) FROM bar;
One raster, several bands
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0), 2, '8BUI', 10, 0), 3, '32BUI', 100, 0) AS rast ) SELECT ST_MapAlgebra( rast, ARRAY[3, 1, 3, 2]::integer[], 'sample_callbackfunc(double precision[], int[], text[])'::regprocedure ) AS rast FROM foo
One raster, one band
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0), 2, '8BUI', 10, 0), 3, '32BUI', 100, 0) AS rast ) SELECT ST_MapAlgebra( rast, 2, 'sample_callbackfunc(double precision[], int[], text[])'::regprocedure ) AS rast FROM foo
Two rasters, two bands
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0), 2, '8BUI', 10, 0), 3, '32BUI', 100, 0) AS rast UNION ALL SELECT 2 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 1, 1, -1, 0, 0, 0), 1, '16BUI', 2, 0), 2, '8BUI', 20, 0), 3, '32BUI', 300, 0) AS rast ) SELECT ST_MapAlgebra( t1.rast, 2, t2.rast, 1, 'sample_callbackfunc(double precision[], int[], text[])'::regprocedure ) AS rast FROM foo t1 CROSS JOIN foo t2 WHERE t1.rid = 1 AND t2.rid = 2
WITH foo AS (SELECT ST_SetBandNoDataValue( ST_SetValue(ST_SetValue(ST_AsRaster( ST_Buffer( ST_GeomFromText('LINESTRING(50 50,100 90,100 50)'), 5,'join=bevel'), 200,200,ARRAY['8BUI'], ARRAY[100], ARRAY[0]), ST_Buffer('POINT(70 70)'::geometry,10,'quad_segs=1') ,50), 'LINESTRING(20 20, 100 100, 150 98)'::geometry,1),0) AS rast ) SELECT 'original' AS title, rast FROM foo UNION ALL SELECT 'no mask mean value' AS title, ST_MapAlgebra(rast,1,'ST_mean4ma(double precision[], int[], text[])'::regprocedure) AS rast FROM foo UNION ALL SELECT 'mask only consider neighbors, exclude center' AS title, ST_MapAlgebra(rast,1,'ST_mean4ma(double precision[], int[], text[])'::regprocedure, '{{1,1,1}, {1,0,1}, {1,1,1}}'::double precision[], false) As rast FROM foo UNION ALL SELECT 'mask weighted only consider neighbors, exclude center multi otehr pixel values by 2' AS title, ST_MapAlgebra(rast,1,'ST_mean4ma(double precision[], int[], text[])'::regprocedure, '{{2,2,2}, {2,0,2}, {2,2,2}}'::double precision[], true) As rast FROM foo;
|
|
|
|
ST_MapAlgebra — Expression version - Returns a one-band raster given one or two input rasters, band indexes and one or more user-specified SQL expressions.
raster ST_MapAlgebra(
raster rast, integer nband, text pixeltype, text expression, double precision nodataval=NULL)
;
raster ST_MapAlgebra(
raster rast, text pixeltype, text expression, double precision nodataval=NULL)
;
raster ST_MapAlgebra(
raster rast1, integer nband1, raster rast2, integer nband2, text expression, text pixeltype=NULL, text extenttype=INTERSECTION, text nodata1expr=NULL, text nodata2expr=NULL, double precision nodatanodataval=NULL)
;
raster ST_MapAlgebra(
raster rast1, raster rast2, text expression, text pixeltype=NULL, text extenttype=INTERSECTION, text nodata1expr=NULL, text nodata2expr=NULL, double precision nodatanodataval=NULL)
;
Expression version - Returns a one-band raster given one or two input rasters, band indexes and one or more user-specified SQL expressions.
Availability: 2.1.0
Creates a new one band raster formed by applying a valid PostgreSQL algebraic operation defined by the expression
on the input raster (rast
). If nband
is not provided, band 1 is assumed. The new raster will have the same georeference, width, and height as the original raster but will only have one band.
If pixeltype
is passed in, then the new raster will have a band of that pixeltype. If pixeltype is passed NULL, then the new raster band will have the same pixeltype as the input rast
band.
Keywords permitted for expression
[rast]
- Pixel value of the pixel of interest
[rast.val]
- Pixel value of the pixel of interest
[rast.x]
- 1-based pixel column of the pixel of interest
[rast.y]
- 1-based pixel row of the pixel of interest
Creates a new one band raster formed by applying a valid PostgreSQL algebraic operation to the two bands defined by the expression
on the two input raster bands rast1
, (rast2
). If no band1
, band2
is specified band 1 is assumed. The resulting raster will be aligned (scale, skew and pixel corners) on the grid defined by the first raster. The resulting raster will have the extent defined by the extenttype
parameter.
A PostgreSQL algebraic expression involving the two rasters and PostgreSQL defined functions/operators that will define the pixel value when pixels intersect. e.g. (([rast1] + [rast2])/2.0)::integer
The resulting pixel type of the output raster. Must be one listed in ST_BandPixelType, left out or set to NULL. If not passed in or set to NULL, will default to the pixeltype of the first raster.
Controls the extent of resulting raster
INTERSECTION
- The extent of the new raster is the intersection of the two rasters. This is the default.
UNION
- The extent of the new raster is the union of the two rasters.
FIRST
- The extent of the new raster is the same as the one of the first raster.
SECOND
- The extent of the new raster is the same as the one of the second raster.
An algebraic expression involving only rast2
or a constant that defines what to return when pixels of rast1
are nodata values and spatially corresponding rast2 pixels have values.
An algebraic expression involving only rast1
or a constant that defines what to return when pixels of rast2
are nodata values and spatially corresponding rast1 pixels have values.
A numeric constant to return when spatially corresponding rast1 and rast2 pixels are both nodata values.
Keywords permitted in expression
, nodata1expr
and nodata2expr
[rast1]
- Pixel value of the pixel of interest from rast1
[rast1.val]
- Pixel value of the pixel of interest from rast1
[rast1.x]
- 1-based pixel column of the pixel of interest from rast1
[rast1.y]
- 1-based pixel row of the pixel of interest from rast1
[rast2]
- Pixel value of the pixel of interest from rast2
[rast2.val]
- Pixel value of the pixel of interest from rast2
[rast2.x]
- 1-based pixel column of the pixel of interest from rast2
[rast2.y]
- 1-based pixel row of the pixel of interest from rast2
WITH foo AS ( SELECT ST_AddBand(ST_MakeEmptyRaster(10, 10, 0, 0, 1, 1, 0, 0, 0), '32BF', 1, -1) AS rast ) SELECT ST_MapAlgebra(rast, 1, NULL, 'ceil([rast]*[rast.x]/[rast.y]+[rast.val])') FROM foo
WITH foo AS ( SELECT 1 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 0, 1, -1, 0, 0, 0), 1, '16BUI', 1, 0), 2, '8BUI', 10, 0), 3, '32BUI', 100, 0) AS rast UNION ALL SELECT 2 AS rid, ST_AddBand(ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(2, 2, 0, 1, 1, -1, 0, 0, 0), 1, '16BUI', 2, 0), 2, '8BUI', 20, 0), 3, '32BUI', 300, 0) AS rast ) SELECT ST_MapAlgebra( t1.rast, 2, t2.rast, 1, '([rast2] + [rast1.val]) / 2' ) AS rast FROM foo t1 CROSS JOIN foo t2 WHERE t1.rid = 1 AND t2.rid = 2
ST_MapAlgebraExpr — 1 raster band version: Creates a new one band raster formed by applying a valid PostgreSQL algebraic operation on the input raster band and of pixeltype provided. Band 1 is assumed if no band is specified.
raster ST_MapAlgebraExpr(
raster rast, integer band, text pixeltype, text expression, double precision nodataval=NULL)
;
raster ST_MapAlgebraExpr(
raster rast, text pixeltype, text expression, double precision nodataval=NULL)
;
ST_MapAlgebraExpr is deprecated as of 2.1.0. Use ST_MapAlgebra instead. |
Creates a new one band raster formed by applying a valid PostgreSQL algebraic operation defined by the expression
on the input raster (rast
). If no band
is specified band 1 is assumed. The new raster will have the same georeference, width, and height as the original raster but will only have one band.
If pixeltype
is passed in, then the new raster will have a band of that pixeltype. If pixeltype is passed NULL, then the new raster band will have the same pixeltype as the input rast
band.
In the expression you can use the term [rast]
to refer to the pixel value of the original band, [rast.x]
to refer to the 1-based pixel column index, [rast.y]
to refer to the 1-based pixel row index.
Disponibilità: 2.0
Create a new 1 band raster from our original that is a function of modulo 2 of the original raster band.
ALTER TABLE dummy_rast ADD COLUMN map_rast raster; UPDATE dummy_rast SET map_rast = ST_MapAlgebraExpr(rast,NULL,'mod([rast],2)') WHERE rid = 2; SELECT ST_Value(rast,1,i,j) As origval, ST_Value(map_rast, 1, i, j) As mapval FROM dummy_rast CROSS JOIN generate_series(1, 3) AS i CROSS JOIN generate_series(1,3) AS j WHERE rid = 2; origval | mapval ---------+-------- 253 | 1 254 | 0 253 | 1 253 | 1 254 | 0 254 | 0 250 | 0 254 | 0 254 | 0
Create a new 1 band raster of pixel-type 2BUI from our original that is reclassified and set the nodata value to be 0.
ALTER TABLE dummy_rast ADD COLUMN map_rast2 raster; UPDATE dummy_rast SET map_rast2 = ST_MapAlgebraExpr(rast,'2BUI','CASE WHEN [rast] BETWEEN 100 and 250 THEN 1 WHEN [rast] = 252 THEN 2 WHEN [rast] BETWEEN 253 and 254 THEN 3 ELSE 0 END', '0') WHERE rid = 2; SELECT DISTINCT ST_Value(rast,1,i,j) As origval, ST_Value(map_rast2, 1, i, j) As mapval FROM dummy_rast CROSS JOIN generate_series(1, 5) AS i CROSS JOIN generate_series(1,5) AS j WHERE rid = 2; origval | mapval ---------+-------- 249 | 1 250 | 1 251 | 252 | 2 253 | 3 254 | 3 SELECT ST_BandPixelType(map_rast2) As b1pixtyp FROM dummy_rast WHERE rid = 2; b1pixtyp ---------- 2BUI
|
|
Create a new 3 band raster same pixel type from our original 3 band raster with first band altered by map algebra and remaining 2 bands unaltered.
SELECT ST_AddBand( ST_AddBand( ST_AddBand( ST_MakeEmptyRaster(rast_view), ST_MapAlgebraExpr(rast_view,1,NULL,'tan([rast])*[rast]') ), ST_Band(rast_view,2) ), ST_Band(rast_view, 3) As rast_view_ma ) FROM wind WHERE rid=167;
ST_MapAlgebraExpr — 2 raster band version: Creates a new one band raster formed by applying a valid PostgreSQL algebraic operation on the two input raster bands and of pixeltype provided. band 1 of each raster is assumed if no band numbers are specified. The resulting raster will be aligned (scale, skew and pixel corners) on the grid defined by the first raster and have its extent defined by the "extenttype" parameter. Values for "extenttype" can be: INTERSECTION, UNION, FIRST, SECOND.
raster ST_MapAlgebraExpr(
raster rast1, raster rast2, text expression, text pixeltype=same_as_rast1_band, text extenttype=INTERSECTION, text nodata1expr=NULL, text nodata2expr=NULL, double precision nodatanodataval=NULL)
;
raster ST_MapAlgebraExpr(
raster rast1, integer band1, raster rast2, integer band2, text expression, text pixeltype=same_as_rast1_band, text extenttype=INTERSECTION, text nodata1expr=NULL, text nodata2expr=NULL, double precision nodatanodataval=NULL)
;
ST_MapAlgebraExpr is deprecated as of 2.1.0. Use ST_MapAlgebra instead. |
Creates a new one band raster formed by applying a valid PostgreSQL algebraic operation to the two bands defined by the expression
on the two input raster bands rast1
, (rast2
). If no band1
, band2
is specified band 1 is assumed. The resulting raster will be aligned (scale, skew and pixel corners) on the grid defined by the first raster. The resulting raster will have the extent defined by the extenttype
parameter.
A PostgreSQL algebraic expression involving the two rasters and PostgreSQL defined functions/operators that will define the pixel value when pixels intersect. e.g. (([rast1] + [rast2])/2.0)::integer
The resulting pixel type of the output raster. Must be one listed in ST_BandPixelType, left out or set to NULL. If not passed in or set to NULL, will default to the pixeltype of the first raster.
Controls the extent of resulting raster
INTERSECTION
- The extent of the new raster is the intersection of the two rasters. This is the default.
UNION
- The extent of the new raster is the union of the two rasters.
FIRST
- The extent of the new raster is the same as the one of the first raster.
SECOND
- The extent of the new raster is the same as the one of the second raster.
An algebraic expression involving only rast2
or a constant that defines what to return when pixels of rast1
are nodata values and spatially corresponding rast2 pixels have values.
An algebraic expression involving only rast1
or a constant that defines what to return when pixels of rast2
are nodata values and spatially corresponding rast1 pixels have values.
A numeric constant to return when spatially corresponding rast1 and rast2 pixels are both nodata values.
If pixeltype
is passed in, then the new raster will have a band of that pixeltype. If pixeltype is passed NULL or no pixel type specified, then the new raster band will have the same pixeltype as the input rast1
band.
Use the term [rast1.val]
[rast2.val]
to refer to the pixel value of the original raster bands and [rast1.x]
, [rast1.y]
etc. to refer to the column / row positions of the pixels.
Disponibilità: 2.0
Create a new 1 band raster from our original that is a function of modulo 2 of the original raster band.
--Create a cool set of rasters -- DROP TABLE IF EXISTS fun_shapes; CREATE TABLE fun_shapes(rid serial PRIMARY KEY, fun_name text, rast raster); -- Insert some cool shapes around Boston in Massachusetts state plane meters -- INSERT INTO fun_shapes(fun_name, rast) VALUES ('ref', ST_AsRaster(ST_MakeEnvelope(235229, 899970, 237229, 901930,26986),200,200,'8BUI',0,0)); INSERT INTO fun_shapes(fun_name,rast) WITH ref(rast) AS (SELECT rast FROM fun_shapes WHERE fun_name = 'ref' ) SELECT 'area' AS fun_name, ST_AsRaster(ST_Buffer(ST_SetSRID(ST_Point(236229, 900930),26986), 1000), ref.rast,'8BUI', 10, 0) As rast FROM ref UNION ALL SELECT 'rand bubbles', ST_AsRaster( (SELECT ST_Collect(geom) FROM (SELECT ST_Buffer(ST_SetSRID(ST_Point(236229 + i*random()*100, 900930 + j*random()*100),26986), random()*20) As geom FROM generate_series(1,10) As i, generate_series(1,10) As j ) As foo ), ref.rast,'8BUI', 200, 0) FROM ref; --map them - SELECT ST_MapAlgebraExpr( area.rast, bub.rast, '[rast2.val]', '8BUI', 'INTERSECTION', '[rast2.val]', '[rast1.val]') As interrast, ST_MapAlgebraExpr( area.rast, bub.rast, '[rast2.val]', '8BUI', 'UNION', '[rast2.val]', '[rast1.val]') As unionrast FROM (SELECT rast FROM fun_shapes WHERE fun_name = 'area') As area CROSS JOIN (SELECT rast FROM fun_shapes WHERE fun_name = 'rand bubbles') As bub
|
|
-- we use ST_AsPNG to render the image so all single band ones look grey -- WITH mygeoms AS ( SELECT 2 As bnum, ST_Buffer(ST_Point(1,5),10) As geom UNION ALL SELECT 3 AS bnum, ST_Buffer(ST_GeomFromText('LINESTRING(50 50,150 150,150 50)'), 10,'join=bevel') As geom UNION ALL SELECT 1 As bnum, ST_Buffer(ST_GeomFromText('LINESTRING(60 50,150 150,150 50)'), 5,'join=bevel') As geom ), -- define our canvas to be 1 to 1 pixel to geometry canvas AS (SELECT ST_AddBand(ST_MakeEmptyRaster(200, 200, ST_XMin(e)::integer, ST_YMax(e)::integer, 1, -1, 0, 0) , '8BUI'::text,0) As rast FROM (SELECT ST_Extent(geom) As e, Max(ST_SRID(geom)) As srid from mygeoms ) As foo ), rbands AS (SELECT ARRAY(SELECT ST_MapAlgebraExpr(canvas.rast, ST_AsRaster(m.geom, canvas.rast, '8BUI', 100), '[rast2.val]', '8BUI', 'FIRST', '[rast2.val]', '[rast1.val]') As rast FROM mygeoms AS m CROSS JOIN canvas ORDER BY m.bnum) As rasts ) SELECT rasts[1] As rast1 , rasts[2] As rast2, rasts[3] As rast3, ST_AddBand( ST_AddBand(rasts[1],rasts[2]), rasts[3]) As final_rast FROM rbands;
|
|
|
|
-- Create new 3 band raster composed of first 2 clipped bands, and overlay of 3rd band with our geometry -- This query took 3.6 seconds on PostGIS windows 64-bit install WITH pr AS -- Note the order of operation: we clip all the rasters to dimensions of our region (SELECT ST_Clip(rast,ST_Expand(geom,50) ) As rast, g.geom FROM aerials.o_2_boston AS r INNER JOIN -- union our parcels of interest so they form a single geometry we can later intersect with (SELECT ST_Union(ST_Transform(the_geom,26986)) AS geom FROM landparcels WHERE pid IN('0303890000', '0303900000')) As g ON ST_Intersects(rast::geometry, ST_Expand(g.geom,50)) ), -- we then union the raster shards together -- ST_Union on raster is kinda of slow but much faster the smaller you can get the rasters -- therefore we want to clip first and then union prunion AS (SELECT ST_AddBand(NULL, ARRAY[ST_Union(rast,1),ST_Union(rast,2),ST_Union(rast,3)] ) As clipped,geom FROM pr GROUP BY geom) -- return our final raster which is the unioned shard with -- with the overlay of our parcel boundaries -- add first 2 bands, then mapalgebra of 3rd band + geometry SELECT ST_AddBand(ST_Band(clipped,ARRAY[1,2]) , ST_MapAlgebraExpr(ST_Band(clipped,3), ST_AsRaster(ST_Buffer(ST_Boundary(geom),2),clipped, '8BUI',250), '[rast2.val]', '8BUI', 'FIRST', '[rast2.val]', '[rast1.val]') ) As rast FROM prunion;
|
ST_MapAlgebraExpr, ST_AddBand, ST_AsPNG, ST_AsRaster, ST_MapAlgebraFct, ST_BandPixelType, ST_GeoReference, ST_Value, ST_Union, ST_Union
ST_MapAlgebraFct — 1 band version - Creates a new one band raster formed by applying a valid PostgreSQL function on the input raster band and of pixeltype prodived. Band 1 is assumed if no band is specified.
raster ST_MapAlgebraFct(
raster rast, regprocedure onerasteruserfunc)
;
raster ST_MapAlgebraFct(
raster rast, regprocedure onerasteruserfunc, text[] VARIADIC args)
;
raster ST_MapAlgebraFct(
raster rast, text pixeltype, regprocedure onerasteruserfunc)
;
raster ST_MapAlgebraFct(
raster rast, text pixeltype, regprocedure onerasteruserfunc, text[] VARIADIC args)
;
raster ST_MapAlgebraFct(
raster rast, integer band, regprocedure onerasteruserfunc)
;
raster ST_MapAlgebraFct(
raster rast, integer band, regprocedure onerasteruserfunc, text[] VARIADIC args)
;
raster ST_MapAlgebraFct(
raster rast, integer band, text pixeltype, regprocedure onerasteruserfunc)
;
raster ST_MapAlgebraFct(
raster rast, integer band, text pixeltype, regprocedure onerasteruserfunc, text[] VARIADIC args)
;
ST_MapAlgebraFct is deprecated as of 2.1.0. Use ST_MapAlgebra instead. |
Creates a new one band raster formed by applying a valid PostgreSQL function specified by the onerasteruserfunc
on the input raster (rast
). If no band
is specified, band 1 is assumed. The new raster will have the same georeference, width, and height as the original raster but will only have one band.
If pixeltype
is passed in, then the new raster will have a band of that pixeltype. If pixeltype is passed NULL, then the new raster band will have the same pixeltype as the input rast
band.
The onerasteruserfunc
parameter must be the name and signature of a SQL or PL/pgSQL function, cast to a regprocedure. A very simple and quite useless PL/pgSQL function example is:
CREATE OR REPLACE FUNCTION simple_function(pixel FLOAT, pos INTEGER[], VARIADIC args TEXT[]) RETURNS FLOAT AS $$ BEGIN RETURN 0.0; END; $$ LANGUAGE 'plpgsql' IMMUTABLE;
The userfunction
may accept two or three arguments: a float value, an optional integer array, and a variadic text array. The first argument is the value of an individual raster cell (regardless of the raster datatype). The second argument is the position of the current processing cell in the form '{x,y}'. The third argument indicates that all remaining parameters to ST_MapAlgebraFct shall be passed through to the userfunction
.
Passing a regprodedure argument to a SQL function requires the full function signature to be passed, then cast to a regprocedure type. To pass the above example PL/pgSQL function as an argument, the SQL for the argument is:
'simple_function(float,integer[],text[])'::regprocedure
Note that the argument contains the name of the function, the types of the function arguments, quotes around the name and argument types, and a cast to a regprocedure.
The third argument to the userfunction
is a variadic text array. All trailing text arguments to any ST_MapAlgebraFct call are passed through to the specified userfunction
, and are contained in the args
argument.
For more information about the VARIADIC keyword, please refer to the PostgreSQL documentation and the "SQL Functions with Variable Numbers of Arguments" section of Query Language (SQL) Functions. |
The text[] argument to the |
Disponibilità: 2.0
Create a new 1 band raster from our original that is a function of modulo 2 of the original raster band.
ALTER TABLE dummy_rast ADD COLUMN map_rast raster; CREATE FUNCTION mod_fct(pixel float, pos integer[], variadic args text[]) RETURNS float AS $$ BEGIN RETURN pixel::integer % 2; END; $$ LANGUAGE 'plpgsql' IMMUTABLE; UPDATE dummy_rast SET map_rast = ST_MapAlgebraFct(rast,NULL,'mod_fct(float,integer[],text[])'::regprocedure) WHERE rid = 2; SELECT ST_Value(rast,1,i,j) As origval, ST_Value(map_rast, 1, i, j) As mapval FROM dummy_rast CROSS JOIN generate_series(1, 3) AS i CROSS JOIN generate_series(1,3) AS j WHERE rid = 2; origval | mapval ---------+-------- 253 | 1 254 | 0 253 | 1 253 | 1 254 | 0 254 | 0 250 | 0 254 | 0 254 | 0
Create a new 1 band raster of pixel-type 2BUI from our original that is reclassified and set the nodata value to a passed parameter to the user function (0).
ALTER TABLE dummy_rast ADD COLUMN map_rast2 raster; CREATE FUNCTION classify_fct(pixel float, pos integer[], variadic args text[]) RETURNS float AS $$ DECLARE nodata float := 0; BEGIN IF NOT args[1] IS NULL THEN nodata := args[1]; END IF; IF pixel < 251 THEN RETURN 1; ELSIF pixel = 252 THEN RETURN 2; ELSIF pixel > 252 THEN RETURN 3; ELSE RETURN nodata; END IF; END; $$ LANGUAGE 'plpgsql'; UPDATE dummy_rast SET map_rast2 = ST_MapAlgebraFct(rast,'2BUI','classify_fct(float,integer[],text[])'::regprocedure, '0') WHERE rid = 2; SELECT DISTINCT ST_Value(rast,1,i,j) As origval, ST_Value(map_rast2, 1, i, j) As mapval FROM dummy_rast CROSS JOIN generate_series(1, 5) AS i CROSS JOIN generate_series(1,5) AS j WHERE rid = 2; origval | mapval ---------+-------- 249 | 1 250 | 1 251 | 252 | 2 253 | 3 254 | 3 SELECT ST_BandPixelType(map_rast2) As b1pixtyp FROM dummy_rast WHERE rid = 2; b1pixtyp ---------- 2BUI
Create a new 3 band raster same pixel type from our original 3 band raster with first band altered by map algebra and remaining 2 bands unaltered.
CREATE FUNCTION rast_plus_tan(pixel float, pos integer[], variadic args text[]) RETURNS float AS $$ BEGIN RETURN tan(pixel) * pixel; END; $$ LANGUAGE 'plpgsql'; SELECT ST_AddBand( ST_AddBand( ST_AddBand( ST_MakeEmptyRaster(rast_view), ST_MapAlgebraFct(rast_view,1,NULL,'rast_plus_tan(float,integer[],text[])'::regprocedure) ), ST_Band(rast_view,2) ), ST_Band(rast_view, 3) As rast_view_ma ) FROM wind WHERE rid=167;
ST_MapAlgebraFct — 2 band version - Creates a new one band raster formed by applying a valid PostgreSQL function on the 2 input raster bands and of pixeltype prodived. Band 1 is assumed if no band is specified. Extent type defaults to INTERSECTION if not specified.
raster ST_MapAlgebraFct(
raster rast1, raster rast2, regprocedure tworastuserfunc, text pixeltype=same_as_rast1, text extenttype=INTERSECTION, text[] VARIADIC userargs)
;
raster ST_MapAlgebraFct(
raster rast1, integer band1, raster rast2, integer band2, regprocedure tworastuserfunc, text pixeltype=same_as_rast1, text extenttype=INTERSECTION, text[] VARIADIC userargs)
;
ST_MapAlgebraFct is deprecated as of 2.1.0. Use ST_MapAlgebra instead. |
Creates a new one band raster formed by applying a valid PostgreSQL function specified by the tworastuserfunc
on the input raster rast1
, rast2
. If no band1
or band2
is specified, band 1 is assumed. The new raster will have the same georeference, width, and height as the original rasters but will only have one band.
If pixeltype
is passed in, then the new raster will have a band of that pixeltype. If pixeltype is passed NULL or left out, then the new raster band will have the same pixeltype as the input rast1
band.
The tworastuserfunc
parameter must be the name and signature of an SQL or PL/pgSQL function, cast to a regprocedure. An example PL/pgSQL function example is:
CREATE OR REPLACE FUNCTION simple_function_for_two_rasters(pixel1 FLOAT, pixel2 FLOAT, pos INTEGER[], VARIADIC args TEXT[]) RETURNS FLOAT AS $$ BEGIN RETURN 0.0; END; $$ LANGUAGE 'plpgsql' IMMUTABLE;
The tworastuserfunc
may accept three or four arguments: a double precision value, a double precision value, an optional integer array, and a variadic text array. The first argument is the value of an individual raster cell in rast1
(regardless of the raster datatype). The second argument is an individual raster cell value in rast2
. The third argument is the position of the current processing cell in the form '{x,y}'. The fourth argument indicates that all remaining parameters to ST_MapAlgebraFct shall be passed through to the tworastuserfunc
.
Passing a regprodedure argument to a SQL function requires the full function signature to be passed, then cast to a regprocedure type. To pass the above example PL/pgSQL function as an argument, the SQL for the argument is:
'simple_function(double precision, double precision, integer[], text[])'::regprocedure
Note that the argument contains the name of the function, the types of the function arguments, quotes around the name and argument types, and a cast to a regprocedure.
The third argument to the tworastuserfunc
is a variadic text array. All trailing text arguments to any ST_MapAlgebraFct call are passed through to the specified tworastuserfunc
, and are contained in the userargs
argument.
For more information about the VARIADIC keyword, please refer to the PostgreSQL documentation and the "SQL Functions with Variable Numbers of Arguments" section of Query Language (SQL) Functions. |
The text[] argument to the |
Disponibilità: 2.0
-- define our user defined function -- CREATE OR REPLACE FUNCTION raster_mapalgebra_union( rast1 double precision, rast2 double precision, pos integer[], VARIADIC userargs text[] ) RETURNS double precision AS $$ DECLARE BEGIN CASE WHEN rast1 IS NOT NULL AND rast2 IS NOT NULL THEN RETURN ((rast1 + rast2)/2.); WHEN rast1 IS NULL AND rast2 IS NULL THEN RETURN NULL; WHEN rast1 IS NULL THEN RETURN rast2; ELSE RETURN rast1; END CASE; RETURN NULL; END; $$ LANGUAGE 'plpgsql' IMMUTABLE COST 1000; -- prep our test table of rasters DROP TABLE IF EXISTS map_shapes; CREATE TABLE map_shapes(rid serial PRIMARY KEY, rast raster, bnum integer, descrip text); INSERT INTO map_shapes(rast,bnum, descrip) WITH mygeoms AS ( SELECT 2 As bnum, ST_Buffer(ST_Point(90,90),30) As geom, 'circle' As descrip UNION ALL SELECT 3 AS bnum, ST_Buffer(ST_GeomFromText('LINESTRING(50 50,150 150,150 50)'), 15) As geom, 'big road' As descrip UNION ALL SELECT 1 As bnum, ST_Translate(ST_Buffer(ST_GeomFromText('LINESTRING(60 50,150 150,150 50)'), 8,'join=bevel'), 10,-6) As geom, 'small road' As descrip ), -- define our canvas to be 1 to 1 pixel to geometry canvas AS ( SELECT ST_AddBand(ST_MakeEmptyRaster(250, 250, ST_XMin(e)::integer, ST_YMax(e)::integer, 1, -1, 0, 0 ) , '8BUI'::text,0) As rast FROM (SELECT ST_Extent(geom) As e, Max(ST_SRID(geom)) As srid from mygeoms ) As foo ) -- return our rasters aligned with our canvas SELECT ST_AsRaster(m.geom, canvas.rast, '8BUI', 240) As rast, bnum, descrip FROM mygeoms AS m CROSS JOIN canvas UNION ALL SELECT canvas.rast, 4, 'canvas' FROM canvas; -- Map algebra on single band rasters and then collect with ST_AddBand INSERT INTO map_shapes(rast,bnum,descrip) SELECT ST_AddBand(ST_AddBand(rasts[1], rasts[2]),rasts[3]), 4, 'map bands overlay fct union (canvas)' FROM (SELECT ARRAY(SELECT ST_MapAlgebraFct(m1.rast, m2.rast, 'raster_mapalgebra_union(double precision, double precision, integer[], text[])'::regprocedure, '8BUI', 'FIRST') FROM map_shapes As m1 CROSS JOIN map_shapes As m2 WHERE m1.descrip = 'canvas' AND m2.descrip <> 'canvas' ORDER BY m2.bnum) As rasts) As foo;
|
CREATE OR REPLACE FUNCTION raster_mapalgebra_userargs( rast1 double precision, rast2 double precision, pos integer[], VARIADIC userargs text[] ) RETURNS double precision AS $$ DECLARE BEGIN CASE WHEN rast1 IS NOT NULL AND rast2 IS NOT NULL THEN RETURN least(userargs[1]::integer,(rast1 + rast2)/2.); WHEN rast1 IS NULL AND rast2 IS NULL THEN RETURN userargs[2]::integer; WHEN rast1 IS NULL THEN RETURN greatest(rast2,random()*userargs[3]::integer)::integer; ELSE RETURN greatest(rast1, random()*userargs[4]::integer)::integer; END CASE; RETURN NULL; END; $$ LANGUAGE 'plpgsql' VOLATILE COST 1000; SELECT ST_MapAlgebraFct(m1.rast, 1, m1.rast, 3, 'raster_mapalgebra_userargs(double precision, double precision, integer[], text[])'::regprocedure, '8BUI', 'INTERSECT', '100','200','200','0') FROM map_shapes As m1 WHERE m1.descrip = 'map bands overlay fct union (canvas)';
ST_MapAlgebraFctNgb — 1-band version: Map Algebra Nearest Neighbor using user-defined PostgreSQL function. Return a raster which values are the result of a PLPGSQL user function involving a neighborhood of values from the input raster band.
raster ST_MapAlgebraFctNgb(
raster rast, integer band, text pixeltype, integer ngbwidth, integer ngbheight, regprocedure onerastngbuserfunc, text nodatamode, text[] VARIADIC args)
;
ST_MapAlgebraFctNgb is deprecated as of 2.1.0. Use ST_MapAlgebra instead. |
(one raster version) Return a raster which values are the result of a PLPGSQL user function involving a neighborhood of values from the input raster band. The user function takes the neighborhood of pixel values as an array of numbers, for each pixel, returns the result from the user function, replacing pixel value of currently inspected pixel with the function result.
Raster on which the user function is evaluated.
Band number of the raster to be evaluated. Default to 1.
The resulting pixel type of the output raster. Must be one listed in ST_BandPixelType or left out or set to NULL. If not passed in or set to NULL, will default to the pixeltype of the rast
. Results are truncated if they are larger than what is allowed for the pixeltype.
La larghezza dell'intorno, in celle.
L'altezza dell'intorno, in celle.
PLPGSQL/psql user function to apply to neighborhood pixels of a single band of a raster. The first element is a 2-dimensional array of numbers representing the rectangular pixel neighborhood
Defines what value to pass to the function for a neighborhood pixel that is nodata or NULL
'ignore': any NODATA values encountered in the neighborhood are ignored by the computation -- this flag must be sent to the user callback function, and the user function decides how to ignore it.
'NULL': any NODATA values encountered in the neighborhood will cause the resulting pixel to be NULL -- the user callback function is skipped in this case.
'value': any NODATA values encountered in the neighborhood are replaced by the reference pixel (the one in the center of the neighborhood). Note that if this value is NODATA, the behavior is the same as 'NULL' (for the affected neighborhood)
argomenti da passare alla funzione utente.
Disponibilità: 2.0
Examples utilize the katrina raster loaded as a single tile described in http://trac.osgeo.org/gdal/wiki/frmts_wtkraster.html and then prepared in the ST_Rescale examples
-- -- A simple 'callback' user function that averages up all the values in a neighborhood. -- CREATE OR REPLACE FUNCTION rast_avg(matrix float[][], nodatamode text, variadic args text[]) RETURNS float AS $$ DECLARE _matrix float[][]; x1 integer; x2 integer; y1 integer; y2 integer; sum float; BEGIN _matrix := matrix; sum := 0; FOR x in array_lower(matrix, 1)..array_upper(matrix, 1) LOOP FOR y in array_lower(matrix, 2)..array_upper(matrix, 2) LOOP sum := sum + _matrix[x][y]; END LOOP; END LOOP; RETURN (sum*1.0/(array_upper(matrix,1)*array_upper(matrix,2) ))::integer ; END; $$ LANGUAGE 'plpgsql' IMMUTABLE COST 1000; -- now we apply to our raster averaging pixels within 2 pixels of each other in X and Y direction -- SELECT ST_MapAlgebraFctNgb(rast, 1, '8BUI', 4,4, 'rast_avg(float[][], text, text[])'::regprocedure, 'NULL', NULL) As nn_with_border FROM katrinas_rescaled limit 1;
|
|
ST_Reclass — Creates a new raster composed of band types reclassified from original. The nband is the band to be changed. If nband is not specified assumed to be 1. All other bands are returned unchanged. Use case: convert a 16BUI band to a 8BUI and so forth for simpler rendering as viewable formats.
raster ST_Reclass(
raster rast, integer nband, text reclassexpr, text pixeltype, double precision nodataval=NULL)
;
raster ST_Reclass(
raster rast, reclassarg[] VARIADIC reclassargset)
;
raster ST_Reclass(
raster rast, text reclassexpr, text pixeltype)
;
Creates a new raster formed by applying a valid PostgreSQL algebraic operation defined by the reclassexpr
on the input raster (rast
). If no band
is specified band 1 is assumed. The new raster will have the same georeference, width, and height as the original raster. Bands not designated will come back unchanged. Refer to reclassarg for description of valid reclassification expressions.
The bands of the new raster will have pixel type of pixeltype
. If reclassargset
is passed in then each reclassarg defines behavior of each band generated.
Disponibilità: 2.0
Create a new raster from the original where band 2 is converted from 8BUI to 4BUI and all values from 101-254 are set to nodata value.
ALTER TABLE dummy_rast ADD COLUMN reclass_rast raster; UPDATE dummy_rast SET reclass_rast = ST_Reclass(rast,2,'0-87:1-10, 88-100:11-15, 101-254:0-0', '4BUI',0) WHERE rid = 2; SELECT i as col, j as row, ST_Value(rast,2,i,j) As origval, ST_Value(reclass_rast, 2, i, j) As reclassval, ST_Value(reclass_rast, 2, i, j, false) As reclassval_include_nodata FROM dummy_rast CROSS JOIN generate_series(1, 3) AS i CROSS JOIN generate_series(1,3) AS j WHERE rid = 2; col | row | origval | reclassval | reclassval_include_nodata -----+-----+---------+------------+--------------------------- 1 | 1 | 78 | 9 | 9 2 | 1 | 98 | 14 | 14 3 | 1 | 122 | | 0 1 | 2 | 96 | 14 | 14 2 | 2 | 118 | | 0 3 | 2 | 180 | | 0 1 | 3 | 99 | 15 | 15 2 | 3 | 112 | | 0 3 | 3 | 169 | | 0
Create a new raster from the original where band 1,2,3 is converted to 1BB,4BUI, 4BUI respectively and reclassified. Note this uses the variadic reclassarg
argument which can take as input an indefinite number of reclassargs (theoretically as many bands as you have)
UPDATE dummy_rast SET reclass_rast = ST_Reclass(rast, ROW(2,'0-87]:1-10, (87-100]:11-15, (101-254]:0-0', '4BUI',NULL)::reclassarg, ROW(1,'0-253]:1, 254:0', '1BB', NULL)::reclassarg, ROW(3,'0-70]:1, (70-86:2, [86-150):3, [150-255:4', '4BUI', NULL)::reclassarg ) WHERE rid = 2; SELECT i as col, j as row,ST_Value(rast,1,i,j) As ov1, ST_Value(reclass_rast, 1, i, j) As rv1, ST_Value(rast,2,i,j) As ov2, ST_Value(reclass_rast, 2, i, j) As rv2, ST_Value(rast,3,i,j) As ov3, ST_Value(reclass_rast, 3, i, j) As rv3 FROM dummy_rast CROSS JOIN generate_series(1, 3) AS i CROSS JOIN generate_series(1,3) AS j WHERE rid = 2; col | row | ov1 | rv1 | ov2 | rv2 | ov3 | rv3 ----+-----+-----+-----+-----+-----+-----+----- 1 | 1 | 253 | 1 | 78 | 9 | 70 | 1 2 | 1 | 254 | 0 | 98 | 14 | 86 | 3 3 | 1 | 253 | 1 | 122 | 0 | 100 | 3 1 | 2 | 253 | 1 | 96 | 14 | 80 | 2 2 | 2 | 254 | 0 | 118 | 0 | 108 | 3 3 | 2 | 254 | 0 | 180 | 0 | 162 | 4 1 | 3 | 250 | 1 | 99 | 15 | 90 | 3 2 | 3 | 254 | 0 | 112 | 0 | 108 | 3 3 | 3 | 254 | 0 | 169 | 0 | 175 | 4
Create a new 3 band (8BUI,8BUI,8BUI viewable raster) from a raster that has only one 32bf band
ALTER TABLE wind ADD COLUMN rast_view raster; UPDATE wind set rast_view = ST_AddBand( NULL, ARRAY[ ST_Reclass(rast, 1,'0.1-10]:1-10,9-10]:11,(11-33:0'::text, '8BUI'::text,0), ST_Reclass(rast,1, '11-33):0-255,[0-32:0,(34-1000:0'::text, '8BUI'::text,0), ST_Reclass(rast,1,'0-32]:0,(32-100:100-255'::text, '8BUI'::text,0) ] );
ST_Union — Returns the union of a set of raster tiles into a single raster composed of 1 or more bands.
raster ST_Union(
setof raster rast)
;
raster ST_Union(
setof raster rast, unionarg[] unionargset)
;
raster ST_Union(
setof raster rast, integer nband)
;
raster ST_Union(
setof raster rast, text uniontype)
;
raster ST_Union(
setof raster rast, integer nband, text uniontype)
;
Returns the union of a set of raster tiles into a single raster composed of at least one band. The resulting raster's extent is the extent of the whole set. In the case of intersection, the resulting value is defined by uniontype
which is one of the following: LAST (default), FIRST, MIN, MAX, COUNT, SUM, MEAN, RANGE.
In order for rasters to be unioned, they most all have the same alignment. Use ST_SameAlignment and ST_NotSameAlignmentReason for more details and help. One way to fix alignment issues is to use ST_Resample and use the same reference raster for alignment. |
Disponibilità: 2.0
Enhanced: 2.1.0 Improved Speed (fully C-Based).
Availability: 2.1.0 ST_Union(rast, unionarg) variant was introduced.
Enhanced: 2.1.0 ST_Union(rast) (variant 1) unions all bands of all input rasters. Prior versions of PostGIS assumed the first band.
Enhanced: 2.1.0 ST_Union(rast, uniontype) (variant 4) unions all bands of all input rasters.
-- this creates a single band from first band of raster tiles -- that form the original file system tile SELECT filename, ST_Union(rast,1) As file_rast FROM sometable WHERE filename IN('dem01', 'dem02') GROUP BY filename;
-- this creates a multi band raster collecting all the tiles that intersect a line -- Note: In 2.0, this would have just returned a single band raster -- , new union works on all bands by default -- this is equivalent to unionarg: ARRAY[ROW(1, 'LAST'), ROW(2, 'LAST'), ROW(3, 'LAST')]::unionarg[] SELECT ST_Union(rast) FROM aerials.boston WHERE ST_Intersects(rast, ST_GeomFromText('LINESTRING(230486 887771, 230500 88772)',26986) );
Here we use the longer syntax if we only wanted a subset of bands or we want to change order of bands
-- this creates a multi band raster collecting all the tiles that intersect a line SELECT ST_Union(rast,ARRAY[ROW(2, 'LAST'), ROW(1, 'LAST'), ROW(3, 'LAST')]::unionarg[]) FROM aerials.boston WHERE ST_Intersects(rast, ST_GeomFromText('LINESTRING(230486 887771, 230500 88772)',26986) );
ST_Distinct4ma — Raster processing function that calculates the number of unique pixel values in a neighborhood.
float8 ST_Distinct4ma(
float8[][] matrix, text nodatamode, text[] VARIADIC args)
;
double precision ST_Distinct4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate the number of unique pixel values in a neighborhood of pixels.
Variant 1 is a specialized callback function for use as a callback parameter to ST_MapAlgebraFctNgb. |
Variant 2 is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Use of Variant 1 is discouraged since ST_MapAlgebraFctNgb has been deprecated as of 2.1.0. |
Disponibilità: 2.0
Enhanced: 2.1.0 Addition of Variant 2
SELECT rid, st_value( st_mapalgebrafctngb(rast, 1, NULL, 1, 1, 'st_distinct4ma(float[][],text,text[])'::regprocedure, 'ignore', NULL), 2, 2 ) FROM dummy_rast WHERE rid = 2; rid | st_value -----+---------- 2 | 3 (1 row)
ST_MapAlgebraFctNgb, ST_MapAlgebra, ST_Min4ma, ST_Max4ma, ST_Sum4ma, ST_Mean4ma, ST_Distinct4ma, ST_StdDev4ma
ST_InvDistWeight4ma — Raster processing function that interpolates a pixel's value from the pixel's neighborhood.
double precision ST_InvDistWeight4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate an interpolated value for a pixel using the Inverse Distance Weighted method.
There are two optional parameters that can be passed through userargs
. The first parameter is the power factor (variable k in the equation below) between 0 and 1 used in the Inverse Distance Weighted equation. If not specified, default value is 1. The second parameter is the weight percentage applied only when the value of the pixel of interest is included with the interpolated value from the neighborhood. If not specified and the pixel of interest has a value, that value is returned.
The basic inverse distance weight equation is:
This function is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Availability: 2.1.0
ST_Max4ma — Raster processing function that calculates the maximum pixel value in a neighborhood.
float8 ST_Max4ma(
float8[][] matrix, text nodatamode, text[] VARIADIC args)
;
double precision ST_Max4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate the maximum pixel value in a neighborhood of pixels.
For Variant 2, a substitution value for NODATA pixels can be specified by passing that value to userargs.
Variant 1 is a specialized callback function for use as a callback parameter to ST_MapAlgebraFctNgb. |
Variant 2 is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Use of Variant 1 is discouraged since ST_MapAlgebraFctNgb has been deprecated as of 2.1.0. |
Disponibilità: 2.0
Enhanced: 2.1.0 Addition of Variant 2
SELECT rid, st_value( st_mapalgebrafctngb(rast, 1, NULL, 1, 1, 'st_max4ma(float[][],text,text[])'::regprocedure, 'ignore', NULL), 2, 2 ) FROM dummy_rast WHERE rid = 2; rid | st_value -----+---------- 2 | 254 (1 row)
ST_MapAlgebraFctNgb, ST_MapAlgebra, ST_Min4ma, ST_Sum4ma, ST_Mean4ma, ST_Range4ma, ST_Distinct4ma, ST_StdDev4ma
ST_Mean4ma — Raster processing function that calculates the mean pixel value in a neighborhood.
float8 ST_Mean4ma(
float8[][] matrix, text nodatamode, text[] VARIADIC args)
;
double precision ST_Mean4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate the mean pixel value in a neighborhood of pixels.
For Variant 2, a substitution value for NODATA pixels can be specified by passing that value to userargs.
Variant 1 is a specialized callback function for use as a callback parameter to ST_MapAlgebraFctNgb. |
Variant 2 is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Use of Variant 1 is discouraged since ST_MapAlgebraFctNgb has been deprecated as of 2.1.0. |
Disponibilità: 2.0
Enhanced: 2.1.0 Addition of Variant 2
SELECT rid, st_value( st_mapalgebrafctngb(rast, 1, '32BF', 1, 1, 'st_mean4ma(float[][],text,text[])'::regprocedure, 'ignore', NULL), 2, 2 ) FROM dummy_rast WHERE rid = 2; rid | st_value -----+------------------ 2 | 253.222229003906 (1 row)
SELECT rid, st_value( ST_MapAlgebra(rast, 1, 'st_mean4ma(double precision[][][], integer[][], text[])'::regprocedure,'32BF', 'FIRST', NULL, 1, 1) , 2, 2) FROM dummy_rast WHERE rid = 2; rid | st_value -----+------------------ 2 | 253.222229003906 (1 row)
ST_MapAlgebraFctNgb, ST_MapAlgebra, ST_Min4ma, ST_Max4ma, ST_Sum4ma, ST_Range4ma, ST_StdDev4ma
ST_Min4ma — Raster processing function that calculates the minimum pixel value in a neighborhood.
float8 ST_Min4ma(
float8[][] matrix, text nodatamode, text[] VARIADIC args)
;
double precision ST_Min4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate the minimum pixel value in a neighborhood of pixels.
For Variant 2, a substitution value for NODATA pixels can be specified by passing that value to userargs.
Variant 1 is a specialized callback function for use as a callback parameter to ST_MapAlgebraFctNgb. |
Variant 2 is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Use of Variant 1 is discouraged since ST_MapAlgebraFctNgb has been deprecated as of 2.1.0. |
Disponibilità: 2.0
Enhanced: 2.1.0 Addition of Variant 2
SELECT rid, st_value( st_mapalgebrafctngb(rast, 1, NULL, 1, 1, 'st_min4ma(float[][],text,text[])'::regprocedure, 'ignore', NULL), 2, 2 ) FROM dummy_rast WHERE rid = 2; rid | st_value -----+---------- 2 | 250 (1 row)
ST_MapAlgebraFctNgb, ST_MapAlgebra, ST_Max4ma, ST_Sum4ma, ST_Mean4ma, ST_Range4ma, ST_Distinct4ma, ST_StdDev4ma
ST_MinDist4ma — Raster processing function that returns the minimum distance (in number of pixels) between the pixel of interest and a neighboring pixel with value.
double precision ST_MinDist4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Return the shortest distance (in number of pixels) between the pixel of interest and the closest pixel with value in the neighborhood.
The intent of this function is to provide an informative data point that helps infer the usefulness of the pixel of interest's interpolated value from ST_InvDistWeight4ma. This function is particularly useful when the neighborhood is sparsely populated. |
This function is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Availability: 2.1.0
ST_Range4ma — Raster processing function that calculates the range of pixel values in a neighborhood.
float8 ST_Range4ma(
float8[][] matrix, text nodatamode, text[] VARIADIC args)
;
double precision ST_Range4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate the range of pixel values in a neighborhood of pixels.
For Variant 2, a substitution value for NODATA pixels can be specified by passing that value to userargs.
Variant 1 is a specialized callback function for use as a callback parameter to ST_MapAlgebraFctNgb. |
Variant 2 is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Use of Variant 1 is discouraged since ST_MapAlgebraFctNgb has been deprecated as of 2.1.0. |
Disponibilità: 2.0
Enhanced: 2.1.0 Addition of Variant 2
SELECT rid, st_value( st_mapalgebrafctngb(rast, 1, NULL, 1, 1, 'st_range4ma(float[][],text,text[])'::regprocedure, 'ignore', NULL), 2, 2 ) FROM dummy_rast WHERE rid = 2; rid | st_value -----+---------- 2 | 4 (1 row)
ST_MapAlgebraFctNgb, ST_MapAlgebra, ST_Min4ma, ST_Max4ma, ST_Sum4ma, ST_Mean4ma, ST_Distinct4ma, ST_StdDev4ma
ST_StdDev4ma — Raster processing function that calculates the standard deviation of pixel values in a neighborhood.
float8 ST_StdDev4ma(
float8[][] matrix, text nodatamode, text[] VARIADIC args)
;
double precision ST_StdDev4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate the standard deviation of pixel values in a neighborhood of pixels.
Variant 1 is a specialized callback function for use as a callback parameter to ST_MapAlgebraFctNgb. |
Variant 2 is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Use of Variant 1 is discouraged since ST_MapAlgebraFctNgb has been deprecated as of 2.1.0. |
Disponibilità: 2.0
Enhanced: 2.1.0 Addition of Variant 2
SELECT rid, st_value( st_mapalgebrafctngb(rast, 1, '32BF', 1, 1, 'st_stddev4ma(float[][],text,text[])'::regprocedure, 'ignore', NULL), 2, 2 ) FROM dummy_rast WHERE rid = 2; rid | st_value -----+------------------ 2 | 1.30170822143555 (1 row)
ST_MapAlgebraFctNgb, ST_MapAlgebra, ST_Min4ma, ST_Max4ma, ST_Sum4ma, ST_Mean4ma, ST_Distinct4ma, ST_StdDev4ma
ST_Sum4ma — Raster processing function that calculates the sum of all pixel values in a neighborhood.
float8 ST_Sum4ma(
float8[][] matrix, text nodatamode, text[] VARIADIC args)
;
double precision ST_Sum4ma(
double precision[][][] value, integer[][] pos, text[] VARIADIC userargs)
;
Calculate the sum of all pixel values in a neighborhood of pixels.
For Variant 2, a substitution value for NODATA pixels can be specified by passing that value to userargs.
Variant 1 is a specialized callback function for use as a callback parameter to ST_MapAlgebraFctNgb. |
Variant 2 is a specialized callback function for use as a callback parameter to ST_MapAlgebra. |
Use of Variant 1 is discouraged since ST_MapAlgebraFctNgb has been deprecated as of 2.1.0. |
Disponibilità: 2.0
Enhanced: 2.1.0 Addition of Variant 2
SELECT rid, st_value( st_mapalgebrafctngb(rast, 1, '32BF', 1, 1, 'st_sum4ma(float[][],text,text[])'::regprocedure, 'ignore', NULL), 2, 2 ) FROM dummy_rast WHERE rid = 2; rid | st_value -----+---------- 2 | 2279 (1 row)
ST_MapAlgebraFctNgb, ST_MapAlgebra, ST_Min4ma, ST_Max4ma, ST_Mean4ma, ST_Range4ma, ST_Distinct4ma, ST_StdDev4ma
ST_Aspect — Returns the aspect (in degrees by default) of an elevation raster band. Useful for analyzing terrain.
raster ST_Aspect(
raster rast, integer band=1, text pixeltype=32BF, text units=DEGREES, boolean interpolate_nodata=FALSE)
;
raster ST_Aspect(
raster rast, integer band, raster customextent, text pixeltype=32BF, text units=DEGREES, boolean interpolate_nodata=FALSE)
;
Returns the aspect (in degrees by default) of an elevation raster band. Utilizes map algebra and applies the aspect equation to neighboring pixels.
units
indicates the units of the aspect. Possible values are: RADIANS, DEGREES (default).
When units
= RADIANS, values are between 0 and 2 * pi radians measured clockwise from North.
When units
= DEGREES, values are between 0 and 360 degrees measured clockwise from North.
If slope of pixel is zero, aspect of pixel is -1.
For more information about Slope, Aspect and Hillshade, please refer to ESRI - How hillshade works and ERDAS Field Guide - Aspect Images. |
Disponibilità: 2.0
Enhanced: 2.1.0 Uses ST_MapAlgebra() and added optional interpolate_nodata
function parameter
Changed: 2.1.0 In prior versions, return values were in radians. Now, return values default to degrees
WITH foo AS ( SELECT ST_SetValues( ST_AddBand(ST_MakeEmptyRaster(5, 5, 0, 0, 1, -1, 0, 0, 0), 1, '32BF', 0, -9999), 1, 1, 1, ARRAY[ [1, 1, 1, 1, 1], [1, 2, 2, 2, 1], [1, 2, 3, 2, 1], [1, 2, 2, 2, 1], [1, 1, 1, 1, 1] ]::double precision[][] ) AS rast ) SELECT ST_DumpValues(ST_Aspect(rast, 1, '32BF')) FROM foo st_dumpvalues ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ ---------------------------------- (1,"{{315,341.565063476562,0,18.4349479675293,45},{288.434936523438,315,0,45,71.5650482177734},{270,270,-1,90,90},{251.565048217773,225,180,135,108.434951782227},{225,198.43495178 2227,180,161.565048217773,135}}") (1 row)
Complete example of tiles of a coverage. This query only works with PostgreSQL 9.1 or higher.
WITH foo AS ( SELECT ST_Tile( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(6, 6, 0, 0, 1, -1, 0, 0, 0), 1, '32BF', 0, -9999 ), 1, 1, 1, ARRAY[ [1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 2, 1], [1, 2, 2, 3, 3, 1], [1, 1, 3, 2, 1, 1], [1, 2, 2, 1, 2, 1], [1, 1, 1, 1, 1, 1] ]::double precision[] ), 2, 2 ) AS rast ) SELECT t1.rast, ST_Aspect(ST_Union(t2.rast), 1, t1.rast) FROM foo t1 CROSS JOIN foo t2 WHERE ST_Intersects(t1.rast, t2.rast) GROUP BY t1.rast;
ST_HillShade — Returns the hypothetical illumination of an elevation raster band using provided azimuth, altitude, brightness and scale inputs.
raster ST_HillShade(
raster rast, integer band=1, text pixeltype=32BF, double precision azimuth=315, double precision altitude=45, double precision max_bright=255, double precision scale=1.0, boolean interpolate_nodata=FALSE)
;
raster ST_HillShade(
raster rast, integer band, raster customextent, text pixeltype=32BF, double precision azimuth=315, double precision altitude=45, double precision max_bright=255, double precision scale=1.0, boolean interpolate_nodata=FALSE)
;
Returns the hypothetical illumination of an elevation raster band using the azimuth, altitude, brightness, and scale inputs. Utilizes map algebra and applies the hill shade equation to neighboring pixels. Return pixel values are between 0 and 255.
azimuth
is a value between 0 and 360 degrees measured clockwise from North.
altitude
is a value between 0 and 90 degrees where 0 degrees is at the horizon and 90 degrees is directly overhead.
max_bright
is a value between 0 and 255 with 0 as no brightness and 255 as max brightness.
scale
is the ratio of vertical units to horizontal. For Feet:LatLon use scale=370400, for Meters:LatLon use scale=111120.
If interpolate_nodata
is TRUE, values for NODATA pixels from the input raster will be interpolated using ST_InvDistWeight4ma before computing the hillshade illumination.
For more information about Hillshade, please refer to How hillshade works. |
Disponibilità: 2.0
Enhanced: 2.1.0 Uses ST_MapAlgebra() and added optional interpolate_nodata
function parameter
Changed: 2.1.0 In prior versions, azimuth and altitude were expressed in radians. Now, azimuth and altitude are expressed in degrees
WITH foo AS ( SELECT ST_SetValues( ST_AddBand(ST_MakeEmptyRaster(5, 5, 0, 0, 1, -1, 0, 0, 0), 1, '32BF', 0, -9999), 1, 1, 1, ARRAY[ [1, 1, 1, 1, 1], [1, 2, 2, 2, 1], [1, 2, 3, 2, 1], [1, 2, 2, 2, 1], [1, 1, 1, 1, 1] ]::double precision[][] ) AS rast ) SELECT ST_DumpValues(ST_Hillshade(rast, 1, '32BF')) FROM foo st_dumpvalues ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ ----------------------------------------------------------------------- (1,"{{NULL,NULL,NULL,NULL,NULL},{NULL,251.32763671875,220.749786376953,147.224319458008,NULL},{NULL,220.749786376953,180.312225341797,67.7497863769531,NULL},{NULL,147.224319458008 ,67.7497863769531,43.1210060119629,NULL},{NULL,NULL,NULL,NULL,NULL}}") (1 row)
Complete example of tiles of a coverage. This query only works with PostgreSQL 9.1 or higher.
WITH foo AS ( SELECT ST_Tile( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(6, 6, 0, 0, 1, -1, 0, 0, 0), 1, '32BF', 0, -9999 ), 1, 1, 1, ARRAY[ [1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 2, 1], [1, 2, 2, 3, 3, 1], [1, 1, 3, 2, 1, 1], [1, 2, 2, 1, 2, 1], [1, 1, 1, 1, 1, 1] ]::double precision[] ), 2, 2 ) AS rast ) SELECT t1.rast, ST_Hillshade(ST_Union(t2.rast), 1, t1.rast) FROM foo t1 CROSS JOIN foo t2 WHERE ST_Intersects(t1.rast, t2.rast) GROUP BY t1.rast;
ST_Roughness — Returns a raster with the calculated "roughness" of a DEM.
raster ST_Roughness(
raster rast, integer nband, raster customextent, text pixeltype="32BF" , boolean interpolate_nodata=FALSE )
;
ST_Slope — Returns the slope (in degrees by default) of an elevation raster band. Useful for analyzing terrain.
raster ST_Slope(
raster rast, integer nband=1, text pixeltype=32BF, text units=DEGREES, double precision scale=1.0, boolean interpolate_nodata=FALSE)
;
raster ST_Slope(
raster rast, integer nband, raster customextent, text pixeltype=32BF, text units=DEGREES, double precision scale=1.0, boolean interpolate_nodata=FALSE)
;
Returns the slope (in degrees by default) of an elevation raster band. Utilizes map algebra and applies the slope equation to neighboring pixels.
units
indicates the units of the slope. Possible values are: RADIANS, DEGREES (default), PERCENT.
scale
is the ratio of vertical units to horizontal. For Feet:LatLon use scale=370400, for Meters:LatLon use scale=111120.
If interpolate_nodata
is TRUE, values for NODATA pixels from the input raster will be interpolated using ST_InvDistWeight4ma before computing the surface slope.
For more information about Slope, Aspect and Hillshade, please refer to ESRI - How hillshade works and ERDAS Field Guide - Slope Images. |
Disponibilità: 2.0
Enhanced: 2.1.0 Uses ST_MapAlgebra() and added optional units
, scale
, interpolate_nodata
function parameters
Changed: 2.1.0 In prior versions, return values were in radians. Now, return values default to degrees
WITH foo AS ( SELECT ST_SetValues( ST_AddBand(ST_MakeEmptyRaster(5, 5, 0, 0, 1, -1, 0, 0, 0), 1, '32BF', 0, -9999), 1, 1, 1, ARRAY[ [1, 1, 1, 1, 1], [1, 2, 2, 2, 1], [1, 2, 3, 2, 1], [1, 2, 2, 2, 1], [1, 1, 1, 1, 1] ]::double precision[][] ) AS rast ) SELECT ST_DumpValues(ST_Slope(rast, 1, '32BF')) FROM foo st_dumpvalues ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ --------------------------------------------------------------------- (1,"{{10.0249881744385,21.5681285858154,26.5650520324707,21.5681285858154,10.0249881744385},{21.5681285858154,35.2643890380859,36.8698959350586,35.2643890380859,21.5681285858154}, {26.5650520324707,36.8698959350586,0,36.8698959350586,26.5650520324707},{21.5681285858154,35.2643890380859,36.8698959350586,35.2643890380859,21.5681285858154},{10.0249881744385,21. 5681285858154,26.5650520324707,21.5681285858154,10.0249881744385}}") (1 row)
Complete example of tiles of a coverage. This query only works with PostgreSQL 9.1 or higher.
WITH foo AS ( SELECT ST_Tile( ST_SetValues( ST_AddBand( ST_MakeEmptyRaster(6, 6, 0, 0, 1, -1, 0, 0, 0), 1, '32BF', 0, -9999 ), 1, 1, 1, ARRAY[ [1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 2, 1], [1, 2, 2, 3, 3, 1], [1, 1, 3, 2, 1, 1], [1, 2, 2, 1, 2, 1], [1, 1, 1, 1, 1, 1] ]::double precision[] ), 2, 2 ) AS rast ) SELECT t1.rast, ST_Slope(ST_Union(t2.rast), 1, t1.rast) FROM foo t1 CROSS JOIN foo t2 WHERE ST_Intersects(t1.rast, t2.rast) GROUP BY t1.rast;
ST_TPI — Returns a raster with the calculated Topographic Position Index.
raster ST_TPI(
raster rast, integer nband, raster customextent, text pixeltype="32BF" , boolean interpolate_nodata=FALSE )
;
ST_TRI — Returns a raster with the calculated Terrain Ruggedness Index.
raster ST_TRI(
raster rast, integer nband, raster customextent, text pixeltype="32BF" , boolean interpolate_nodata=FALSE )
;
Box3D — Returns the box 3d representation of the enclosing box of the raster.
box3d Box3D(
raster rast)
;
Returns the box representing the extent of the raster.
The polygon is defined by the corner points of the bounding box ((MINX
, MINY
), (MAXX
, MAXY
))
Changed: 2.0.0 In pre-2.0 versions, there used to be a box2d instead of box3d. Since box2d is a deprecated type, this was changed to box3d.
ST_ConvexHull — Return the convex hull geometry of the raster including pixel values equal to BandNoDataValue. For regular shaped and non-skewed rasters, this gives the same result as ST_Envelope so only useful for irregularly shaped or skewed rasters.
geometry ST_ConvexHull(
raster rast)
;
Return the convex hull geometry of the raster including the NoDataBandValue band pixels. For regular shaped and non-skewed rasters, this gives more or less the same result as ST_Envelope so only useful for irregularly shaped or skewed rasters.
ST_Envelope floors the coordinates and hence add a little buffer around the raster so the answer is subtly different from ST_ConvexHull which does not floor. |
Refer to PostGIS Raster Specification for a diagram of this.
-- Note envelope and convexhull are more or less the same SELECT ST_AsText(ST_ConvexHull(rast)) As convhull, ST_AsText(ST_Envelope(rast)) As env FROM dummy_rast WHERE rid=1; convhull | env --------------------------------------------------------+------------------------------------ POLYGON((0.5 0.5,20.5 0.5,20.5 60.5,0.5 60.5,0.5 0.5)) | POLYGON((0 0,20 0,20 60,0 60,0 0))
-- now we skew the raster -- note how the convex hull and envelope are now different SELECT ST_AsText(ST_ConvexHull(rast)) As convhull, ST_AsText(ST_Envelope(rast)) As env FROM (SELECT ST_SetRotation(rast, 0.1, 0.1) As rast FROM dummy_rast WHERE rid=1) As foo; convhull | env --------------------------------------------------------+------------------------------------ POLYGON((0.5 0.5,20.5 1.5,22.5 61.5,2.5 60.5,0.5 0.5)) | POLYGON((0 0,22 0,22 61,0 61,0 0))
ST_DumpAsPolygons — Returns a set of geomval (geom,val) rows, from a given raster band. If no band number is specified, band num defaults to 1.
setof geomval ST_DumpAsPolygons(
raster rast, integer band_num=1, boolean exclude_nodata_value=TRUE)
;
This is a set-returning function (SRF). It returns a set of geomval rows, formed by a geometry (geom) and a pixel band value (val). Each polygon is the union of all pixels for that band that have the same pixel value denoted by val.
ST_DumpAsPolygon is useful for polygonizing rasters. It is the reverse of a GROUP BY in that it creates new rows. For example it can be used to expand a single raster into multiple POLYGONS/MULTIPOLYGONS.
Availability: Requires GDAL 1.7 or higher.
If there is a no data value set for a band, pixels with that value will not be returned. |
If you only care about count of pixels with a given value in a raster, it is faster to use ST_ValueCount. |
This is different than ST_PixelAsPolygons where one geometry is returned for each pixel regardless of pixel value. |
SELECT val, ST_AsText(geom) As geomwkt FROM ( SELECT (ST_DumpAsPolygons(rast)).* FROM dummy_rast WHERE rid = 2 ) As foo WHERE val BETWEEN 249 and 251 ORDER BY val; val | geomwkt -----+-------------------------------------------------------------------------- 249 | POLYGON((3427927.95 5793243.95,3427927.95 5793243.85,3427928 5793243.85, 3427928 5793243.95,3427927.95 5793243.95)) 250 | POLYGON((3427927.75 5793243.9,3427927.75 5793243.85,3427927.8 5793243.85, 3427927.8 5793243.9,3427927.75 5793243.9)) 250 | POLYGON((3427927.8 5793243.8,3427927.8 5793243.75,3427927.85 5793243.75, 3427927.85 5793243.8, 3427927.8 5793243.8)) 251 | POLYGON((3427927.75 5793243.85,3427927.75 5793243.8,3427927.8 5793243.8, 3427927.8 5793243.85,3427927.75 5793243.85))
ST_Envelope — Returns the polygon representation of the extent of the raster.
geometry ST_Envelope(
raster rast)
;
Returns the polygon representation of the extent of the raster in spatial coordinate units defined by srid. It is a float8 minimum bounding box represented as a polygon.
The polygon is defined by the corner points of the bounding box ((MINX
, MINY
), (MINX
, MAXY
), (MAXX
, MAXY
), (MAXX
, MINY
), (MINX
, MINY
))
SELECT rid, ST_AsText(ST_Envelope(rast)) As envgeomwkt FROM dummy_rast; rid | envgeomwkt -----+-------------------------------------------------------------------- 1 | POLYGON((0 0,20 0,20 60,0 60,0 0)) 2 | POLYGON((3427927 5793243,3427928 5793243, 3427928 5793244,3427927 5793244, 3427927 5793243))
ST_MinConvexHull — Return the convex hull geometry of the raster excluding NODATA pixels.
geometry ST_MinConvexHull(
raster rast, integer nband=NULL)
;
Return the convex hull geometry of the raster excluding NODATA pixels. If nband
is NULL, all bands of the raster are considered.
Availability: 2.1.0
WITH foo AS ( SELECT ST_SetValues( ST_SetValues( ST_AddBand(ST_AddBand(ST_MakeEmptyRaster(9, 9, 0, 0, 1, -1, 0, 0, 0), 1, '8BUI', 0, 0), 2, '8BUI', 1, 0), 1, 1, 1, ARRAY[ [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 1, 0, 0, 0, 0, 1], [0, 0, 0, 1, 1, 0, 0, 0, 0], [0, 0, 0, 1, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0] ]::double precision[][] ), 2, 1, 1, ARRAY[ [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 0, 0, 0, 0, 1, 0, 0, 0], [0, 0, 0, 0, 1, 1, 0, 0, 0], [0, 0, 0, 0, 0, 1, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 1, 0, 0, 0, 0, 0, 0] ]::double precision[][] ) AS rast ) SELECT ST_AsText(ST_ConvexHull(rast)) AS hull, ST_AsText(ST_MinConvexHull(rast)) AS mhull, ST_AsText(ST_MinConvexHull(rast, 1)) AS mhull_1, ST_AsText(ST_MinConvexHull(rast, 2)) AS mhull_2 FROM foo hull | mhull | mhull_1 | mhull_2 ----------------------------------+-------------------------------------+-------------------------------------+------------------------------------- POLYGON((0 0,9 0,9 -9,0 -9,0 0)) | POLYGON((0 -3,9 -3,9 -9,0 -9,0 -3)) | POLYGON((3 -3,9 -3,9 -6,3 -6,3 -3)) | POLYGON((0 -3,6 -3,6 -9,0 -9,0 -3))
ST_Polygon — Returns a multipolygon geometry formed by the union of pixels that have a pixel value that is not no data value. If no band number is specified, band num defaults to 1.
geometry ST_Polygon(
raster rast, integer band_num=1)
;
Availability: 0.1.6 Requires GDAL 1.7 or higher.
Enhanced: 2.1.0 Improved Speed (fully C-Based) and the returning multipolygon is ensured to be valid.
Changed: 2.1.0 In prior versions would sometimes return a polygon, changed to always return multipolygon.
-- by default no data band value is 0 or not set, so polygon will return a square polygon SELECT ST_AsText(ST_Polygon(rast)) As geomwkt FROM dummy_rast WHERE rid = 2; geomwkt -------------------------------------------- MULTIPOLYGON(((3427927.75 5793244,3427928 5793244,3427928 5793243.75,3427927.75 5793243.75,3427927.75 5793244))) -- now we change the no data value of first band UPDATE dummy_rast SET rast = ST_SetBandNoDataValue(rast,1,254) WHERE rid = 2; SELECt rid, ST_BandNoDataValue(rast) from dummy_rast where rid = 2; -- ST_Polygon excludes the pixel value 254 and returns a multipolygon SELECT ST_AsText(ST_Polygon(rast)) As geomwkt FROM dummy_rast WHERE rid = 2; geomwkt --------------------------------------------------------- MULTIPOLYGON(((3427927.9 5793243.95,3427927.85 5793243.95,3427927.85 5793244,3427927.9 5793244,3427927.9 5793243.95)),((3427928 5793243.85,3427928 5793243.8,3427927.95 5793243.8,3427927.95 5793243.85,3427927.9 5793243.85,3427927.9 5793243.9,3427927.9 5793243.95,3427927.95 5793243.95,3427928 5793243.95,3427928 5793243.85)),((3427927.8 5793243.75,3427927.75 5793243.75,3427927.75 5793243.8,3427927.75 5793243.85,3427927.75 5793243.9,3427927.75 5793244,3427927.8 5793244,3427927.8 5793243.9,3427927.8 5793243.85,3427927.85 5793243.85,3427927.85 5793243.8,3427927.85 5793243.75,3427927.8 5793243.75))) -- Or if you want the no data value different for just one time SELECT ST_AsText( ST_Polygon( ST_SetBandNoDataValue(rast,1,252) ) ) As geomwkt FROM dummy_rast WHERE rid =2; geomwkt --------------------------------- MULTIPOLYGON(((3427928 5793243.85,3427928 5793243.8,3427928 5793243.75,3427927.85 5793243.75,3427927.8 5793243.75,3427927.8 5793243.8,3427927.75 5793243.8,3427927.75 5793243.85,3427927.75 5793243.9,3427927.75 5793244,3427927.8 5793244,3427927.85 5793244,3427927.9 5793244,3427928 5793244,3427928 5793243.95,3427928 5793243.85),(3427927.9 5793243.9,3427927.9 5793243.85,3427927.95 5793243.85,3427927.95 5793243.9,3427927.9 5793243.9)))
TRUE
if A's bounding box intersects B's bounding box.TRUE
if A's bounding box is to the left of B's.TRUE
if A's bounding box is to the right of B's.TRUE
if A's bounding box is the same as B's. Uses double precision bounding box.TRUE
if A's bounding box is contained by B's. Uses double precision bounding box.TRUE
if A's bounding box is the same as B's.TRUE
if A's bounding box is contains B's. Uses double precision bounding box.&& — Returns TRUE
if A's bounding box intersects B's bounding box.
boolean &&(
raster A , raster B )
;
boolean &&(
raster A , geometry B )
;
boolean &&(
geometry B , raster A )
;
&< — Returns TRUE
if A's bounding box is to the left of B's.
boolean &<(
raster A , raster B )
;
&> — Returns TRUE
if A's bounding box is to the right of B's.
boolean &>(
raster A , raster B )
;
= — Returns TRUE
if A's bounding box is the same as B's. Uses double precision bounding box.
boolean =(
raster A , raster B )
;
The =
operator returns TRUE
if the bounding box of raster A is the same as the bounding box of raster B. PostgreSQL uses the =, <, and > operators defined for rasters to perform internal orderings and comparison of rasters (ie. in a GROUP BY or ORDER BY clause).
This operand will NOT make use of any indexes that may be available on the rasters. Use ~= instead. This operator exists mostly so one can group by the raster column. |
Availability: 2.1.0
@ — Returns TRUE
if A's bounding box is contained by B's. Uses double precision bounding box.
boolean @(
raster A , raster B )
;
boolean @(
geometry A , raster B )
;
boolean @(
raster B , geometry A )
;
~= — Returns TRUE
if A's bounding box is the same as B's.
boolean ~=(
raster A , raster B )
;
The ~=
operator returns TRUE
if the bounding box of raster A is the same as the bounding box of raster B.
This operand will make use of any indexes that may be available on the rasters. |
Disponibilità: 2.0
~ — Returns TRUE
if A's bounding box is contains B's. Uses double precision bounding box.
boolean ~(
raster A , raster B )
;
boolean ~(
geometry A , raster B )
;
boolean ~(
raster B , geometry A )
;
ST_Contains — Return true if no points of raster rastB lie in the exterior of raster rastA and at least one point of the interior of rastB lies in the interior of rastA.
boolean ST_Contains(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_Contains(
raster rastA , raster rastB )
;
Raster rastA contains rastB if and only if no points of rastB lie in the exterior of rastA and at least one point of the interior of rastB lies in the interior of rastA. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This function will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_Contains(ST_Polygon(raster), geometry) or ST_Contains(geometry, ST_Polygon(raster)). |
ST_Contains() is the inverse of ST_Within(). So, ST_Contains(rastA, rastB) implies ST_Within(rastB, rastA). |
Availability: 2.1.0
-- specified band numbers SELECT r1.rid, r2.rid, ST_Contains(r1.rast, 1, r2.rast, 1) FROM dummy_rast r1 CROSS JOIN dummy_rast r2 WHERE r1.rid = 1; NOTICE: The first raster provided has no bands rid | rid | st_contains -----+-----+------------- 1 | 1 | 1 | 2 | f
-- no band numbers specified SELECT r1.rid, r2.rid, ST_Contains(r1.rast, r2.rast) FROM dummy_rast r1 CROSS JOIN dummy_rast r2 WHERE r1.rid = 1; rid | rid | st_contains -----+-----+------------- 1 | 1 | t 1 | 2 | f
ST_ContainsProperly — Return true if rastB intersects the interior of rastA but not the boundary or exterior of rastA.
boolean ST_ContainsProperly(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_ContainsProperly(
raster rastA , raster rastB )
;
Raster rastA contains properly rastB if rastB intersects the interior of rastA but not the boundary or exterior of rastA. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
Raster rastA does not contain properly itself but does contain itself.
This function will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_ContainsProperly(ST_Polygon(raster), geometry) or ST_ContainsProperly(geometry, ST_Polygon(raster)). |
Availability: 2.1.0
ST_Covers — Return true if no points of raster rastB lie outside raster rastA.
boolean ST_Covers(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_Covers(
raster rastA , raster rastB )
;
Raster rastA covers rastB if and only if no points of rastB lie in the exterior of rastA. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This function will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_Covers(ST_Polygon(raster), geometry) or ST_Covers(geometry, ST_Polygon(raster)). |
Availability: 2.1.0
ST_CoveredBy — Return true if no points of raster rastA lie outside raster rastB.
boolean ST_CoveredBy(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_CoveredBy(
raster rastA , raster rastB )
;
Raster rastA is covered by rastB if and only if no points of rastA lie in the exterior of rastB. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This function will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_CoveredBy(ST_Polygon(raster), geometry) or ST_CoveredBy(geometry, ST_Polygon(raster)). |
Availability: 2.1.0
ST_Disjoint — Return true if raster rastA does not spatially intersect rastB.
boolean ST_Disjoint(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_Disjoint(
raster rastA , raster rastB )
;
Raster rastA and rastB are disjointed if they do not share any space together. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This function does NOT use any indexes. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_Disjoint(ST_Polygon(raster), geometry). |
Availability: 2.1.0
-- rid = 1 has no bands, hence the NOTICE and the NULL value for st_disjoint SELECT r1.rid, r2.rid, ST_Disjoint(r1.rast, 1, r2.rast, 1) FROM dummy_rast r1 CROSS JOIN dummy_rast r2 WHERE r1.rid = 2; NOTICE: The second raster provided has no bands rid | rid | st_disjoint -----+-----+------------- 2 | 1 | 2 | 2 | f
-- this time, without specifying band numbers SELECT r1.rid, r2.rid, ST_Disjoint(r1.rast, r2.rast) FROM dummy_rast r1 CROSS JOIN dummy_rast r2 WHERE r1.rid = 2; rid | rid | st_disjoint -----+-----+------------- 2 | 1 | t 2 | 2 | f
ST_Intersects — Return true if raster rastA spatially intersects raster rastB.
boolean ST_Intersects(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_Intersects(
raster rastA , raster rastB )
;
boolean ST_Intersects(
raster rast , integer nband , geometry geommin )
;
boolean ST_Intersects(
raster rast , geometry geommin , integer nband=NULL )
;
boolean ST_Intersects(
geometry geommin , raster rast , integer nband=NULL )
;
Return true if raster rastA spatially intersects raster rastB. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This function will make use of any indexes that may be available on the rasters. |
Enhanced: 2.0.0 support raster/raster intersects was introduced.
Changed: 2.1.0 The behavior of the ST_Intersects(raster, geometry) variants changed to match that of ST_Intersects(geometry, raster). |
ST_Overlaps — Return true if raster rastA and rastB intersect but one does not completely contain the other.
boolean ST_Overlaps(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_Overlaps(
raster rastA , raster rastB )
;
Return true if raster rastA spatially overlaps raster rastB. This means that rastA and rastB intersect but one does not completely contain the other. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This function will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_Overlaps(ST_Polygon(raster), geometry). |
Availability: 2.1.0
ST_Touches — Return true if raster rastA and rastB have at least one point in common but their interiors do not intersect.
boolean ST_Touches(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_Touches(
raster rastA , raster rastB )
;
Return true if raster rastA spatially touches raster rastB. This means that rastA and rastB have at least one point in common but their interiors do not intersect. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This function will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_Touches(ST_Polygon(raster), geometry). |
Availability: 2.1.0
ST_SameAlignment — Returns true if rasters have same skew, scale, spatial ref, and offset (pixels can be put on same grid without cutting into pixels) and false if they don't with notice detailing issue.
boolean ST_SameAlignment(
raster rastA , raster rastB )
;
boolean ST_SameAlignment(
double precision ulx1 , double precision uly1 , double precision scalex1 , double precision scaley1 , double precision skewx1 , double precision skewy1 , double precision ulx2 , double precision uly2 , double precision scalex2 , double precision scaley2 , double precision skewx2 , double precision skewy2 )
;
boolean ST_SameAlignment(
raster set rastfield )
;
Non-Aggregate version (Variants 1 and 2): Returns true if the two rasters (either provided directly or made using the values for upperleft, scale, skew and srid) have the same scale, skew, srid and at least one of any of the four corners of any pixel of one raster falls on any corner of the grid of the other raster. Returns false if they don't and a NOTICE detailing the alignment issue.
Aggregate version (Variant 3): From a set of rasters, returns true if all rasters in the set are aligned. The ST_SameAlignment() function is an "aggregate" function in the terminology of PostgreSQL. That means that it operates on rows of data, in the same way the SUM() and AVG() functions do.
Disponibilità: 2.0
Enhanced: 2.1.0 addition of Aggegrate variant
SELECT ST_SameAlignment( ST_MakeEmptyRaster(1, 1, 0, 0, 1, 1, 0, 0), ST_MakeEmptyRaster(1, 1, 0, 0, 1, 1, 0, 0) ) as sm; sm ---- t
SELECT ST_SameAlignment(A.rast,b.rast) FROM dummy_rast AS A CROSS JOIN dummy_rast AS B; NOTA: I due raster forniti hanno SRID differenti st_samealignment ------------------ t f f f
ST_NotSameAlignmentReason — Returns text stating if rasters are aligned and if not aligned, a reason why.
text ST_NotSameAlignmentReason(
raster rastA, raster rastB)
;
Returns text stating if rasters are aligned and if not aligned, a reason why.
If there are several reasons why the rasters are not aligned, only one reason (the first test to fail) will be returned. |
Availability: 2.1.0
SELECT ST_SameAlignment( ST_MakeEmptyRaster(1, 1, 0, 0, 1, 1, 0, 0), ST_MakeEmptyRaster(1, 1, 0, 0, 1.1, 1.1, 0, 0) ), ST_NotSameAlignmentReason( ST_MakeEmptyRaster(1, 1, 0, 0, 1, 1, 0, 0), ST_MakeEmptyRaster(1, 1, 0, 0, 1.1, 1.1, 0, 0) ) ; st_samealignment | st_notsamealignmentreason ------------------+------------------------------------------------- f | The rasters have different scales on the X axis (1 row)
ST_Within — Return true if no points of raster rastA lie in the exterior of raster rastB and at least one point of the interior of rastA lies in the interior of rastB.
boolean ST_Within(
raster rastA , integer nbandA , raster rastB , integer nbandB )
;
boolean ST_Within(
raster rastA , raster rastB )
;
Raster rastA is within rastB if and only if no points of rastA lie in the exterior of rastB and at least one point of the interior of rastA lies in the interior of rastB. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
This operand will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_Within(ST_Polygon(raster), geometry) or ST_Within(geometry, ST_Polygon(raster)). |
ST_Within() is the inverse of ST_Contains(). So, ST_Within(rastA, rastB) implies ST_Contains(rastB, rastA). |
Availability: 2.1.0
ST_DWithin — Return true if rasters rastA and rastB are within the specified distance of each other.
boolean ST_DWithin(
raster rastA , integer nbandA , raster rastB , integer nbandB , double precision distance_of_srid )
;
boolean ST_DWithin(
raster rastA , raster rastB , double precision distance_of_srid )
;
Return true if rasters rastA and rastB are within the specified distance of each other. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
The distance is specified in units defined by the spatial reference system of the rasters. For this function to make sense, the source rasters must both be of the same coordinate projection, having the same SRID.
This operand will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_DWithin(ST_Polygon(raster), geometry). |
Availability: 2.1.0
ST_DFullyWithin — Return true if rasters rastA and rastB are fully within the specified distance of each other.
boolean ST_DFullyWithin(
raster rastA , integer nbandA , raster rastB , integer nbandB , double precision distance_of_srid )
;
boolean ST_DFullyWithin(
raster rastA , raster rastB , double precision distance_of_srid )
;
Return true if rasters rastA and rastB are fully within the specified distance of each other. If the band number is not provided (or set to NULL), only the convex hull of the raster is considered in the test. If the band number is provided, only those pixels with value (not NODATA) are considered in the test.
The distance is specified in units defined by the spatial reference system of the rasters. For this function to make sense, the source rasters must both be of the same coordinate projection, having the same SRID.
This operand will make use of any indexes that may be available on the rasters. |
To test the spatial relationship of a raster and a geometry, use ST_Polygon on the raster, e.g. ST_DFullyWithin(ST_Polygon(raster), geometry). |
Availability: 2.1.0
10.1. | I'm getting error | |||
As of PostGIS 2.1.3 and 2.0.5, a security change was made to by default disable all GDAL drivers and out of db rasters. The release notes are at PostGIS 2.0.6, 2.1.3 security release. In order to reenable specific drivers or all drivers and reenable out of database support, refer to Section 2.1, “Versione sintetica”. | ||||
10.2. | Dove posso trovare altre informazioni sul progetto PostGIS Raster? | |||
Visita la home page di PostGIS Raster. | ||||
10.3. | Ci sono libri o manuali che mi possano introdurre a questa meravigliosa invenzione? | |||
C'è un tutorial completo per principianti, dal titolo Intersecting vector buffers with large raster coverage using PostGIS Raster. Jorge ha nel suo blog una serie di articoli su PostGIS Raster che mostrano come caricare dati raster e che danno un confronto con le funzioni analoghe di Oracle GeoRaster. Controlla la serie di articoli di Jorge su PostGIS Raster / Oracle GeoRaster. C'è un intero capitolo (più di 35 pagine) dedicato a PostGIS Raster, assieme a codice libero e dati da scaricare alla pagina PostGIS in Action - Raster chapter. Potete acquistare PostGIS in Action da Manning in versione cartacea (con sconti significativi per grossi quantitativi) oppore in versione E-book. Potete anche acquistare il libro da Amazon e da altri distributori. Tutte le copie cartacee contengono un buono per scaricare la versione E-book. Qui trovi una rassegna da parte di un utente di PostGIS Raster per applicazioni di classificazione del suolo nelle zone verdi urbane. | ||||
10.4. | Come installo il supporto Raster nel mio database PostGIS? | |||
Il modo più semplice è scaricare la versione binaria di PostGIS e della parte Raster, che sono attualmente disponibili per Windows e per le ultime versioni di Max OSX, partendo da una versione funzionante di PostGIS 2.0.0 o versioni superiori, con PostgreSQL 8.4, 9.0, o 9.1. Nota che in PostGIS 2.0 la parte Raster è completamente integrata, per cui dovrebbe essere compilata quando compilate PostGIS. Le istruzioni per installare ed eseguire il programma sotto Windows sono disponibili alla pagina How to Install and Configure PostGIS raster on windows If you are on windows, you can compile yourself, or use the pre-compiled PostGIS Raster windows binaries. If you are on Mac OSX Leopard or Snow Leopard, there are binaries available at Kyng Chaos Mac OSX PostgreSQL/GIS binaries. Poi per abilitare il supporto raster nel tuo database, esegui il file rtpostgis.sql nel tuo database. Per aggiornare un'installazione esistente, utilizza rtpostgis_upgrade_minor..sql anziché rtpostgis.sql Per le altre piattaforme, in genere dovrai eseguire la compilazione da solo. Ci sono dipendenze da PostGISe da GDAL. Per ulteriori dettagli sulla compilazione da sorgente, vedi Installing PostGIS Raster from source (in prior versions of PostGIS) | ||||
10.5. | Ottengo l'errore could not load library "C:/Program Files/PostgreSQL/8.4/lib/rtpostgis.dll": The specified module could not be found. or could not load library on Linux when trying to run rtpostgis.sql | |||
rtpostgis.so/dll is viene compilato con la dipendenza da libgdal.dll/so. Assicurati su Windows di avere libgdal-1.dll nella cartellabin folder della tua installazione PostgreSQL. Per Linux, libgdal deve essere nel tuo path o nella cartella bin. Potresti anche riscontrare errori diversi se non hai PostGIS installato nel tuo database. Assicurati di installare PostGIS prima di provare a installare il supporto raster. | ||||
10.6. | Come carico i dati raster in PostGIS? | |||
La più recente versione di PostGIS è corredata dall'eseguibile | ||||
10.7. | Che tipo di formati raster posso caricare nel mio database? | |||
Qualsiasi formato supportato dalla tua libreria GDAL. I formati supportati da GDAL sono documentati. La tua particolare versione di GDAL potrebbe non supportare tutti i formati. Per verificare di preciso, puoi usare raster2pgsql -G | ||||
10.8. | Posso esportare i miei dati raster PostGIS ad altri formati raster? | |||
Sì Dalla versione 1.7 in su, GDAL ha un driver per PostGIS Raster, ma viene compilato solo se scegli la compilazione con supporto PostgreSQL. Al momento il driver non supporta i raster a blocchi irregolari, sebbene sia possibile memorizzare blocchi irregolari nel tipo dati PostGIS raster. Se stai compilando il programma a partire dal sorgente, devi includere nel sorgente l'opzione --with-pg=path/to/pg_config per abilitare il driver. Ti rimandiamo alla pagina GDAL Build Hints per suggerimenti sulla compilazione di GDAL su divrse piattaforme. Se la tua versione di GDAL è compilata con il driver PostGIS Raster, dovresti trovare PostGIS Raster nella lista quando esegui gdalinfo --formats Per ottenere informazioni riassuntive sul tuo raster tramite GDAL, utilizza gdalinfo: gdalinfo "PG:host=localhost port=5432 dbname='mygisdb' user='postgres' password='whatever' schema='someschema' table=sometable"
Per esportare dati verso altri formati raster, utilizza gdal_translate. L'esempio sotto esporterà tutti i dati da una tabella a un file PNG col 10% della dimensione. A seconda dei tipi di banda dei pixel, alcune conversioni non funzioneranno, se il formato di esportazione non supprta un dato tipo di pixel. Per esempio, i tipi di banda a virgola mobile, o gli interi senza segno a 32 bit non si convertiranno bene in JPEG. Ecco un esempio di una semplice conversione gdal_translate -of PNG -outsize 10% 10% "PG:host=localhost port=5432 dbname='mygisdb' user='postgres' password='whatever' schema='someschema' table=sometable" C:\somefile.png Puoi anche utilizzare la clausola SQL WHERE nell'esportazione, riportando il testo where=.... nella stringa di collegamento del driver. Sotto si riportano alcuni esempi che utilizzano la clausola where gdal_translate -of PNG -outsize 10% 10% "PG:host=localhost port=5432 dbname='mygisdb' user='postgres' password='whatever' schema='someschema' table=sometable where='filename=\'abcd.sid\''" " C:\somefile.png gdal_translate -of PNG -outsize 10% 10% "PG:host=localhost port=5432 dbname='mygisdb' user='postgres' password='whatever' schema='someschema' table=sometable where='ST_Intersects(rast, ST_SetSRID(ST_Point(-71.032,42.3793),4326) )' " C:\intersectregion.png Per vedere altri esempi e sintassi, si rimanda al paragrafo Leggere dati raster da un raster PostGIS | ||||
10.9. | I binari di GDAL sono già disponibili con il supporto raster PostGIS? | |||
Sì. Controlla la pagina GDAL Binaries. Qualsiasi versione compilata con l'opzione di supporto per PostgreSQL dovrebbe avere il supporto per i raster PostGIS. PostGIS Raster is undergoing many changes. If you want to get the latest nightly build for Windows -- then check out the Tamas Szekeres nightly builds built with Visual Studio which contain GDAL trunk, Python Bindings and MapServer executables and PostGIS Raster driver built-in. Just click the SDK bat and run your commands from there. http://www.gisinternals.com. Also available are VS project files. La versione più recente e stabile di FWTools per Windows è compilata con il supporto raster. | ||||
10.10. | Quali strumenti posso usare per visualizzare i dati raster PostGIS? | |||
Puoi utilizzare Mapserver compilato con GDAL 1.7 o versioni superiori e con il supporto del driver raster PostGIS per vedere i dati raster. Quantum GIS (QGIS) inoltre supporta la visualizzazione dei raster PostGIS se hai il driver installato. In teoria qualsiasi strumento che utilizzi GDAL per la restituzione dei dati può supportare i raster PostGIS direttamente o con un minimo sforzo. Sempre per Windows, I binari di Tama http://vbkto.dyndns.org/sdk/ sono una buona scelta se non volete prendervi la briga di eseguire la compilazione in proprio. | ||||
10.11. | Come posso aggiungere uno strato raster PostGIS in una mappa di MapServer? | |||
Anzitutto ti serve GDAL 1.7 o superiore compilato con il supporto per raster PostGIS. E consigliata la versione 1.8 o superiore, dato che quest'ultima risolve numerosi problemi, e altri problemi con i raster PostGIS sono risolti nella versione trunk. You can much like you can with any other raster. Refer to MapServer Raster processing options for list of various processing functions you can use with MapServer raster layers. What makes PostGIS raster data particularly interesting, is that since each tile can have various standard database columns, you can segment it in your data source Sotto trovi un esempio di come definire uno strato raster PostGIS in MapServer.
-- visualizzare un raster con le opzioni standard LAYER NAME coolwktraster TYPE raster STATUS ON DATA "PG:host=localhost port=5432 dbname='somedb' user='someuser' password='whatever' schema='someschema' table='cooltable' mode='2'" PROCESSING "NODATA=0" PROCESSING "SCALE=AUTO" #... other standard raster processing functions here #... classes are optional but useful for 1 band data CLASS NAME "boring" EXPRESSION ([pixel] < 20) COLOR 250 250 250 END CLASS NAME "mildly interesting" EXPRESSION ([pixel] > 20 AND [pixel] < 1000) COLOR 255 0 0 END CLASS NAME "very interesting" EXPRESSION ([pixel] >= 1000) COLOR 0 255 0 END END -- visualizzare un raster con le opzioni standard e la clausola where LAYER NAME soil_survey2009 TYPE raster STATUS ON DATA "PG:host=localhost port=5432 dbname='somedb' user='someuser' password='whatever' schema='someschema' table='cooltable' where='survey_year=2009' mode='2'" PROCESSING "NODATA=0" #... other standard raster processing functions here #... classes are optional but useful for 1 band data END | ||||
10.12. | Quali funzioni posso attualmente utilizzare con i dati raster? | |||
Ti rimandiamo alla lista Chapter 9, Raster Reference. Ce ne sono anche altre, ma si tratta di funzioni in via di sviluppo. Ti rimandiamo alla pagina PostGIS Raster roadmap page per i dettagli sugli sviluppi attesti in futuro. | ||||
10.13. | Ottengo un errore ERROR: function st_intersects(raster, unknown) is not unique or st_union(geometry,text) is not unique. Come lo aggiusto? | |||
L'errore "function is not unique" si ha quando uno dei vostri argomenti e la rappresentazione testuale di una geometria, anziché una geometria vera e propria. In questi casi, PostgreSQL segna la rappresentazione testuale come tipo sconosciuto, che potrebbe essere applicato sia a st_intersects(raster, geometry) che a st_intersects(raster,raster), portando così a una situazione non univoca, dato che entrambe le funzioni supportano la tua richiesta. Al fine di evitare questa situazione, devi eseguire il cast della geometria in modo che sia effettivamente interpretata come geometria. Per esempio, se il tuo codice è tipo: SELECT rast FROM my_raster WHERE ST_Intersects(rast, 'SRID=4326;POINT(-10 10)'); Dei eseguire il cast della geometria come testo a una geometria vera e propria in questo modo: SELECT rast FROM my_raster WHERE ST_Intersects(rast, 'SRID=4326;POINT(-10 10)'::geometry); | ||||
10.14. | Quali sono le differenze tra i raster di PostGIS, i GeoRaster di Oracle (SDO_GEORASTER) e i tipi SDO_RASTER? | |||
Per una discussione dettagliata su questo argomento, dai un'occhiata all'articolo di Jorge Arévalo Oracle GeoRaster and PostGIS Raster: First impressions The major advantage of one-georeference-by-raster over one-georeference-by-layer is to allow: * coverages to be not necessarily rectangular (which is often the case of raster coverage covering large extents. See the possible raster arrangements in the documentation) * rasters to overlaps (which is necessary to implement lossless vector to raster conversion) These arrangements are possible in Oracle as well, but they imply the storage of multiple SDO_GEORASTER objects linked to as many SDO_RASTER tables. A complex coverage can lead to hundreds of tables in the database. With PostGIS Raster you can store a similar raster arrangement into a unique table. It's a bit like if PostGIS would force you to store only full rectangular vector coverage without gaps or overlaps (a perfect rectangular topological layer). This is very practical in some applications but practice has shown that it is not realistic or desirable for most geographical coverages. Vector structures needs the flexibility to store discontinuous and non-rectangular coverages. We think it is a big advantage that raster structure should benefit as well. | ||||
10.15. | raster2pgsql load of large file fails with String of N bytes is too long for encoding conversion? | |||
raster2pgsql doesn't make any connections to your database when generating the file to load. If your database has set an explicit client encoding different from your database encoding, then when loading large raster files (above 30 MB in size), you may run into a This generally happens if for example you have your database in UTF8, but to support windows apps, you have the client encoding set to To work around this make sure the client encoding is the same as your database encoding during load. You can do this by explicitly setting the encoding in your load script. Example, if you are on windows: set PGCLIENTENCODING=UTF8 If you are on Unix/Linux export PGCLIENTENCODING=UTF8 Gory details of this issue are detailed in http://trac.osgeo.org/postgis/ticket/2209 |
Table of Contents
I tipi e le funzioni della Topologia PostGIS si usano per gestire oggetti topologici quali facce, bordi e nodi.
La presentazione di Sandro Santilli alla conferenza "PostGIS Day Paris 2011" da' una buona sintesi sulla Topologia PostGIS e i futuri sviluppi Topology with PostGIS 2.0 slide deck.
Vincent Picavet provides a good synopsis and overview of what is Topology, how is it used, and various FOSS4G tools that support it in PostGIS Topology PGConf EU 2012.
Un esempio di database GIS topologico e' il US Census Topologically Integrated Geographic Encoding and Reference System (TIGER) database. Se vuoi sperimentare con la topologia PostGIS e hai bisogno di dati, dai un'occhiata a Topology_Load_Tiger.
Il modulo topologico era presente in versioni precedenti di PostGIS ma non era mai stato parte della documentazione ufficiale. A partire da PostGIS 2.0.0 e' in corso una ripulitura generale per rimuovere le funzioni deprecate, risolvere noti problemi di usabilita', migliorare la documentazione delle funzioni, aggiungere funzioni e aumentare la conformita' con gli standard SQL-MM.
Dettagli del progetto si possono trovare sul PostGIS Topology Wiki.
Tutte le funzioni e le tabelle associate con questo modulo sono installate in uno schema chiamato topology
.
Le funzioni definite dallo standard SQL/MM hanno il prefisso ST_ e le funzioni specifiche di PostGIS sono senza prefisso.
Per abilitare la topologia in PostGIS 2.0, configura con il parametro --with-topology come descritto in Chapter 2, Installazione PostGIS. Alcune funzioni dipendono da GEOS 3.3+ quindi dovresti compilare con GEOS 3.3+ per utilizzare a pieno il supporto topologico.
ST_GetFaceEdges
.ValidateTopology
getfaceedges_returntype — Un tipo composito che consiste in un numero sequenziale e un numero di "bordo". E' il tipo ritornato dalla funzione ST_GetFaceEdges
.
Un tipo composito che consiste in un numero sequenziale e un numero di "bordo". E' il tipo ritornato dalla funzione ST_GetFaceEdges
.
sequence
e' un intero: si riferisce ad una topologia definita nella tabella topology.topology, che definisce lo schema e lo srid per la topologia.
edge
e' un intero: l'identificativo di un bordo.
TopoGeometry — A composite type representing a topologically defined geometry
A composite type that refers to a topology geometry in a specific topology layer, having a specific type and a specific id. The elements of a TopoGeometry are the properties: topology_id, layer_id, id integer, type integer.
topology_id
is an integer: Refers to a topology defined in the topology.topology table which defines the topology schema and srid.
layer_id
is an integer: The layer_id in the layers table that the TopoGeometry belongs to. The combination of topology_id, layer_id provides a unique reference in the topology.layers table.
id
is an integer: The id is the autogenerated sequence number that uniquely defines the topogeometry in the respective topology layer.
type
integer between 1 - 4 that defines the geometry type: 1:[multi]point, 2:[multi]line, 3:[multi]poly, 4:collection
validatetopology_returntype — A composite type that consists of an error message and id1 and id2 to denote location of error. This is the return type for ValidateTopology
A composite type that consists of an error message and two integers. The ValidateTopology function returns a set of these to denote validation errors and the id1 and id2 to denote the ids of the topology objects involved in the error.
error
is varchar: Denotes type of error.
Current error descriptors are: coincident nodes, edge crosses node, edge not simple, edge end node geometry mis-match, edge start node geometry mismatch, face overlaps face,face within face,
id1
is an integer: Denotes identifier of edge / face / nodes in error.
id2
is an integer: For errors that involve 2 objects denotes the secondary edge / or node
Abstract
This section lists the PostgreSQL domains installed by PostGIS Topology. Domains can be used like object types as return objects of functions or table columns. The distinction between a domain and a type is that a domain is an existing type with a check constraint bound to it.
TopoElement — An array of 2 integers generally used to identify a TopoGeometry component.
An array of 2 integers used to represent one component of a simple or hierarchical TopoGeometry.
In the case of a simple TopoGeometry the first element of the array represents the identifier of a topological primitive and the second element represents its type (1:node, 2:edge, 3:face). In the case of a hierarchical TopoGeometry the first element of the array represents the identifier of a child TopoGeometry and the second element represents its layer identifier.
For any given hierarchical TopoGeometry all child TopoGeometry elements will come from the same child layer, as specified in the topology.layer record for the layer of the TopoGeometry being defined. |
SELECT te[1] AS id, te[2] AS type FROM ( SELECT ARRAY[1,2]::topology.topoelement AS te ) f; id | type ----+------ 1 | 2
SELECT ARRAY[1,2]::topology.topoelement; te ------- {1,2}
--Example of what happens when you try to case a 3 element array to topoelement -- NOTE: topoement has to be a 2 element array so fails dimension check SELECT ARRAY[1,2,3]::topology.topoelement; ERROR: value for domain topology.topoelement violates check constraint "dimensions"
TopoElementArray — An array of TopoElement objects
An array of 1 or more TopoElement objects, generally used to pass around components of TopoGeometry objects.
SELECT '{{1,2},{4,3}}'::topology.topoelementarray As tea; tea ------- {{1,2},{4,3}} -- more verbose equivalent -- SELECT ARRAY[ARRAY[1,2], ARRAY[4,3]]::topology.topoelementarray As tea; tea ------- {{1,2},{4,3}} --using the array agg function packaged with topology -- SELECT topology.TopoElementArray_Agg(ARRAY[e,t]) As tea FROM generate_series(1,4) As e CROSS JOIN generate_series(1,3) As t; tea -------------------------------------------------------------------------- {{1,1},{1,2},{1,3},{2,1},{2,2},{2,3},{3,1},{3,2},{3,3},{4,1},{4,2},{4,3}}
SELECT '{{1,2,4},{3,4,5}}'::topology.topoelementarray As tea; ERROR: value for domain topology.topoelementarray violates check constraint "dimensions"
table_name
in schema schema_name
and unregisters the columns from topology.layer table.AddTopoGeometryColumn — Adds a topogeometry column to an existing table, registers this new column as a layer in topology.layer and returns the new layer_id.
text AddTopoGeometryColumn(
varchar topology_name, varchar schema_name, varchar table_name, varchar column_name, varchar feature_type)
;
text AddTopoGeometryColumn(
varchar topology_name, varchar schema_name, varchar table_name, varchar column_name, varchar feature_type, integer child_layer)
;
Each TopoGeometry object belongs to a specific Layer of a specific Topology. Before creating a TopoGeometry object you need to create its TopologyLayer. A Topology Layer is an association of a feature-table with the topology. It also contain type and hierarchy information. We create a layer using the AddTopoGeometryColumn() function:
This function will both add the requested column to the table and add a record to the topology.layer table with all the given info.
If you don't specify [child_layer] (or set it to NULL) this layer would contain Basic TopoGeometries (composed by primitive topology elements). Otherwise this layer will contain hierarchical TopoGeometries (composed by TopoGeometries from the child_layer).
Once the layer is created (its id is returned by the AddTopoGeometryColumn function) you're ready to construct TopoGeometry objects in it
Valid feature_type
s are: POINT, LINE, POLYGON, COLLECTION
Availability: 1.?
-- Note for this example we created our new table in the ma_topo schema -- though we could have created it in a different schema -- in which case topology_name and schema_name would be different CREATE SCHEMA ma; CREATE TABLE ma.parcels(gid serial, parcel_id varchar(20) PRIMARY KEY, address text); SELECT topology.AddTopoGeometryColumn('ma_topo', 'ma', 'parcels', 'topo', 'POLYGON');
CREATE SCHEMA ri; CREATE TABLE ri.roads(gid serial PRIMARY KEY, road_name text); SELECT topology.AddTopoGeometryColumn('ri_topo', 'ri', 'roads', 'topo', 'LINE');
DropTopology — Use with caution: Drops a topology schema and deletes its reference from topology.topology table and references to tables in that schema from the geometry_columns table.
integer DropTopology(
varchar topology_schema_name)
;
Drops a topology schema and deletes its reference from topology.topology table and references to tables in that schema from the geometry_columns table. This function should be USED WITH CAUTION, as it could destroy data you care about. If the schema does not exist, it just removes reference entries the named schema.
Availability: 1.?
DropTopoGeometryColumn — Drops the topogeometry column from the table named table_name
in schema schema_name
and unregisters the columns from topology.layer table.
text DropTopoGeometryColumn(
varchar schema_name, varchar table_name, varchar column_name)
;
TopologySummary — Takes a topology name and provides summary totals of types of objects in topology
text TopologySummary(
varchar topology_schema_name)
;
Takes a topology name and provides summary totals of types of objects in topology.
Availability: 2.0.0
SELECT topology.topologysummary('city_data'); topologysummary -------------------------------------------------------- Topology city_data (329), SRID 4326, precision: 0 22 nodes, 24 edges, 10 faces, 29 topogeoms in 5 layers Layer 1, type Polygonal (3), 9 topogeoms Deploy: features.land_parcels.feature Layer 2, type Puntal (1), 8 topogeoms Deploy: features.traffic_signs.feature Layer 3, type Lineal (2), 8 topogeoms Deploy: features.city_streets.feature Layer 4, type Polygonal (3), 3 topogeoms Hierarchy level 1, child layer 1 Deploy: features.big_parcels.feature Layer 5, type Puntal (1), 1 topogeoms Hierarchy level 1, child layer 2 Deploy: features.big_signs.feature
ValidateTopology — Returns a set of validatetopology_returntype objects detailing issues with topology
setof validatetopology_returntype ValidateTopology(
varchar topology_schema_name)
;
Returns a set of validatetopology_returntype objects detailing issues with topology. List of possible errors and what the returned ids represent are displayed below:
Error | id1 | id2 |
---|---|---|
edge crosses node | edge_id | node_id |
invalid edge | edge_id | null |
edge not simple | edge_id | null |
edge crosses edge | edge_id | edge_id |
edge start node geometry mis-match | edge_id | node_id |
edge end node geometry mis-match | edge_id | node_id |
face without edges | face_id | null |
face has no rings | face_id | null |
face overlaps face | face_id | face_id |
face within face | inner face_id | outer face_id |
Availability: 1.0.0
Enhanced: 2.0.0 more efficient edge crossing detection and fixes for false positives that were existent in prior versions.
Changed: 2.2.0 values for id1 and id2 were swapped for 'edge crosses node' to be consistent with error description.
CreateTopology — Creates a new topology schema and registers this new schema in the topology.topology table.
integer CreateTopology(
varchar topology_schema_name)
;
integer CreateTopology(
varchar topology_schema_name, integer srid)
;
integer CreateTopology(
varchar topology_schema_name, integer srid, double precision tolerance)
;
integer CreateTopology(
varchar topology_schema_name, integer srid, double precision tolerance, boolean hasz)
;
Creates a new schema with name topology_name
consisting of tables (edge_data
,face
,node
, relation
and registers this new topology in the topology.topology table. It returns the id of the topology in the topology table. The srid is the spatial reference identified as defined in spatial_ref_sys table for that topology. Topologies must be uniquely named. The tolerance is measured in the units of the spatial reference system. If the tolerance is not specified defaults to 0.
This is similar to the SQL/MM ST_InitTopoGeo but a bit more functional. hasz
defaults to false if not specified.
Availability: 1.?
This example creates a new schema called ma_topo that will store edges, faces, and relations in Massachusetts State Plane meters. The tolerance represents 1/2 meter since the spatial reference system is a meter based spatial reference system
SELECT topology.CreateTopology('ma_topo',26986, 0.5);
Create Rhode Island topology in State Plane ft
SELECT topology.CreateTopology('ri_topo',3438) As topoid; topoid ------ 2
CopyTopology — Makes a copy of a topology structure (nodes, edges, faces, layers and TopoGeometries).
integer CopyTopology(
varchar existing_topology_name, varchar new_name)
;
Creates a new topology with name new_topology_name
and SRID and precision taken from existing_topology_name
, copies all nodes, edges and faces in there, copies layers and their TopoGeometries too.
The new rows in topology.layer will contain synthesized values for schema_name, table_name and feature_column. This is because the TopoGeometry will only exist as a definition but won't be available in any user-level table yet. |
Availability: 2.0.0
ST_InitTopoGeo — Creates a new topology schema and registers this new schema in the topology.topology table and details summary of process.
text ST_InitTopoGeo(
varchar topology_schema_name)
;
This is an SQL-MM equivalent of CreateTopology but lacks the spatial reference and tolerance options of CreateTopology and outputs a text description of creation instead of topology id.
Availability: 1.?
This method implements the SQL/MM specification. SQL-MM 3 Topo-Geo and Topo-Net 3: Routine Details: X.3.17
ST_CreateTopoGeo — Adds a collection of geometries to a given empty topology and returns a message detailing success.
text ST_CreateTopoGeo(
varchar atopology, geometry acollection)
;
Adds a collection of geometries to a given empty topology and returns a message detailing success.
Useful for populating an empty topology.
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details -- X.3.18
-- Populate topology -- SELECT topology.ST_CreateTopoGeo('ri_topo', ST_GeomFromText('MULTILINESTRING((384744 236928,384750 236923,384769 236911,384799 236895,384811 236890,384833 236884, 384844 236882,384866 236881,384879 236883,384954 236898,385087 236932,385117 236938, 385167 236938,385203 236941,385224 236946,385233 236950,385241 236956,385254 236971, 385260 236979,385268 236999,385273 237018,385273 237037,385271 237047,385267 237057, 385225 237125,385210 237144,385192 237161,385167 237192,385162 237202,385159 237214, 385159 237227,385162 237241,385166 237256,385196 237324,385209 237345,385234 237375, 385237 237383,385238 237399,385236 237407,385227 237419,385213 237430,385193 237439, 385174 237451,385170 237455,385169 237460,385171 237475,385181 237503,385190 237521, 385200 237533,385206 237538,385213 237541,385221 237542,385235 237540,385242 237541, 385249 237544,385260 237555,385270 237570,385289 237584,385292 237589,385291 237596,385284 237630))',3438) ); st_createtopogeo ---------------------------- Topology ri_topo populated -- create tables and topo geometries -- CREATE TABLE ri.roads(gid serial PRIMARY KEY, road_name text); SELECT topology.AddTopoGeometryColumn('ri_topo', 'ri', 'roads', 'topo', 'LINE');
TopoGeo_AddPoint — Adds a point to an existing topology using a tolerance and possibly splitting an existing edge.
integer TopoGeo_AddPoint(
varchar toponame, geometry apoint, float8 tolerance)
;
TopoGeo_AddLineString — Adds a linestring to an existing topology using a tolerance and possibly splitting existing edges/faces. Returns edge identifiers
SETOF integer TopoGeo_AddLineString(
varchar toponame, geometry aline, float8 tolerance)
;
TopoGeo_AddPolygon — Adds a polygon to an existing topology using a tolerance and possibly splitting existing edges/faces.
integer TopoGeo_AddPolygon(
varchar atopology, geometry apoly, float8 atolerance)
;
alinestring
to a topology connecting two existing isolated nodes anode
and anothernode
and returns the edge id of the new edge.apoint
geometry exists as a node an error is thrown. REturns description of move.ST_AddIsoNode — Adds an isolated node to a face in a topology and returns the nodeid of the new node. If face is null, the node is still created.
integer ST_AddIsoNode(
varchar atopology, integer aface, geometry apoint)
;
Adds an isolated node with point location apoint
to an existing face with faceid aface
to a topology atopology
and returns the nodeid of the new node.
If the spatial reference system (srid) of the point geometry is not the same as the topology, the apoint
is not a point geometry, the point is null, or the point intersects an existing edge (even at the boundaries) then an exception is thrown. If the point already exists as a node, an exception is thrown.
If aface
is not null and the apoint
is not within the face, then an exception is thrown.
Availability: 1.?
This method implements the SQL/MM specification. SQL-MM: Topo-Net Routines: X+1.3.1
ST_AddIsoEdge — Adds an isolated edge defined by geometry alinestring
to a topology connecting two existing isolated nodes anode
and anothernode
and returns the edge id of the new edge.
integer ST_AddIsoEdge(
varchar atopology, integer anode, integer anothernode, geometry alinestring)
;
Adds an isolated edge defined by geometry alinestring
to a topology connecting two existing isolated nodes anode
and anothernode
and returns the edge id of the new edge.
If the spatial reference system (srid) of the alinestring
geometry is not the same as the topology, any of the input arguments are null, or the nodes are contained in more than one face, or the nodes are start or end nodes of an existing edge, then an exception is thrown.
If the alinestring
is not within the face of the face the anode
and anothernode
belong to, then an exception is thrown.
If the anode
and anothernode
are not the start and end points of the alinestring
then an exception is thrown.
Availability: 1.?
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.4
ST_AddEdgeNewFaces — Add a new edge and, if in doing so it splits a face, delete the original face and replace it with two new faces.
integer ST_AddEdgeNewFaces(
varchar atopology, integer anode, integer anothernode, geometry acurve)
;
Add a new edge and, if in doing so it splits a face, delete the original face and replace it with two new faces. Returns the id of the newly added edge.
Updates all existing joined edges and relationships accordingly.
If any arguments are null, the given nodes are unknown (must already exist in the node
table of the topology schema) , the acurve
is not a LINESTRING
, the anode
and anothernode
are not the start and endpoints of acurve
then an error is thrown.
If the spatial reference system (srid) of the acurve
geometry is not the same as the topology an exception is thrown.
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.12
ST_AddEdgeModFace — Add a new edge and, if in doing so it splits a face, modify the original face and add a new face.
integer ST_AddEdgeModFace(
varchar atopology, integer anode, integer anothernode, geometry acurve)
;
Add a new edge and, if doing so splits a face, modify the original face and add a new one.
If possible, the new face will be created on left side of the new edge. This will not be possible if the face on the left side will need to be the Universe face (unbounded). |
Returns the id of the newly added edge.
Updates all existing joined edges and relationships accordingly.
If any arguments are null, the given nodes are unknown (must already exist in the node
table of the topology schema) , the acurve
is not a LINESTRING
, the anode
and anothernode
are not the start and endpoints of acurve
then an error is thrown.
If the spatial reference system (srid) of the acurve
geometry is not the same as the topology an exception is thrown.
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.13
ST_RemEdgeNewFace — Removes an edge and, if the removed edge separated two faces, delete the original faces and replace them with a new face.
integer ST_RemEdgeNewFace(
varchar atopology, integer anedge)
;
Removes an edge and, if the removed edge separated two faces, delete the original faces and replace them with a new face.
Returns the id of a newly created face or NULL, if no new face is created. No new face is created when the removed edge is dangling or isolated or confined with the universe face (possibly making the universe flood into the face on the other side).
Updates all existing joined edges and relationships accordingly.
Refuses to remove an edge partecipating in the definition of an existing TopoGeometry. Refuses to heal two faces if any TopoGeometry is defined by only one of them (and not the other).
If any arguments are null, the given edge is unknown (must already exist in the edge
table of the topology schema), the topology name is invalid then an error is thrown.
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.14
ST_RemEdgeModFace — Removes an edge and, if the removed edge separated two faces, delete one of the them and modify the other to take the space of both.
integer ST_RemEdgeModFace(
varchar atopology, integer anedge)
;
Removes an edge and, if the removed edge separated two faces, delete one of the them and modify the other to take the space of both. Preferentially keeps the face on the right, to be symmetric with ST_AddEdgeModFace also keeping it. Returns the id of the face remaining in place of the removed edge.
Updates all existing joined edges and relationships accordingly.
Refuses to remove an edge partecipating in the definition of an existing TopoGeometry. Refuses to heal two faces if any TopoGeometry is defined by only one of them (and not the other).
If any arguments are null, the given edge is unknown (must already exist in the edge
table of the topology schema), the topology name is invalid then an error is thrown.
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.15
ST_ChangeEdgeGeom — Changes the shape of an edge without affecting the topology structure.
integer ST_ChangeEdgeGeom(
varchar atopology, integer anedge, geometry acurve)
;
Changes the shape of an edge without affecting the topology structure.
If any arguments are null, the given edge does not exist in the node
table of the topology schema, the acurve
is not a LINESTRING
, the anode
and anothernode
are not the start and endpoints of acurve
or the modification would change the underlying topology then an error is thrown.
If the spatial reference system (srid) of the acurve
geometry is not the same as the topology an exception is thrown.
If the new acurve
is not simple, then an error is thrown.
If moving the edge from old to new position would hit an obstacle then an error is thrown.
Disponibilità: 1.1.0
Enhanced: 2.0.0 adds topological consistency enforcement
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details X.3.6
ST_ModEdgeSplit — Split an edge by creating a new node along an existing edge, modifying the original edge and adding a new edge.
integer ST_ModEdgeSplit(
varchar atopology, integer anedge, geometry apoint)
;
Split an edge by creating a new node along an existing edge, modifying the original edge and adding a new edge. Updates all existing joined edges and relationships accordingly. Returns the identifier of the newly added node.
Availability: 1.?
Changed: 2.0 - In prior versions, this was misnamed ST_ModEdgesSplit
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.9
-- Add an edge -- SELECT topology.AddEdge('ma_topo', ST_GeomFromText('LINESTRING(227592 893910, 227600 893910)', 26986) ) As edgeid; -- edgeid- 3 -- Split the edge -- SELECT topology.ST_ModEdgeSplit('ma_topo', 3, ST_SetSRID(ST_Point(227594,893910),26986) ) As node_id; node_id ------------------------- 7
ST_ModEdgeHeal — Heal two edges by deleting the node connecting them, modifying the first edge and deleting the second edge. Returns the id of the deleted node.
int ST_ModEdgeHeal(
varchar atopology, integer anedge, integer anotheredge)
;
Heal two edges by deleting the node connecting them, modifying the first edge and deleting the second edge. Returns the id of the deleted node. Updates all existing joined edges and relationships accordingly.
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.9
ST_NewEdgeHeal — Heal two edges by deleting the node connecting them, deleting both edges, and replacing them with an edge whose direction is the same as the first edge provided.
int ST_NewEdgeHeal(
varchar atopology, integer anedge, integer anotheredge)
;
Heal two edges by deleting the node connecting them, deleting both edges, and replacing them with an edge whose direction is the same as the first edge provided. Returns the id of the new edge replacing the healed ones. Updates all existing joined edges and relationships accordingly.
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X.3.9
ST_MoveIsoNode — Moves an isolated node in a topology from one point to another. If new apoint
geometry exists as a node an error is thrown. REturns description of move.
text ST_MoveIsoNode(
varchar atopology, integer anedge, geometry apoint)
;
Moves an isolated node in a topology from one point to another. If new apoint
geometry exists as a node an error is thrown.
If any arguments are null, the apoint
is not a point, the existing node is not isolated (is a start or end point of an existing edge), new node location intersects an existing edge (even at the end points) then an exception is thrown.
If the spatial reference system (srid) of the point geometry is not the same as the topology an exception is thrown.
Availability: 1.?
This method implements the SQL/MM specification. SQL-MM: Topo-Net Routines: X.3.2
-- Add an isolated node with no face -- SELECT topology.ST_AddIsoNode('ma_topo', NULL, ST_GeomFromText('POINT(227579 893916)', 26986) ) As nodeid; nodeid -------- 7 -- Move the new node -- SELECT topology.ST_MoveIsoNode('ma_topo', 7, ST_GeomFromText('POINT(227579.5 893916.5)', 26986) ) As descrip; descrip ---------------------------------------------------- Isolated Node 7 moved to location 227579.5,893916.5
ST_NewEdgesSplit — Split an edge by creating a new node along an existing edge, deleting the original edge and replacing it with two new edges. Returns the id of the new node created that joins the new edges.
integer ST_NewEdgesSplit(
varchar atopology, integer anedge, geometry apoint)
;
Split an edge with edge id anedge
by creating a new node with point location apoint
along current edge, deleting the original edge and replacing it with two new edges. Returns the id of the new node created that joins the new edges. Updates all existing joined edges and relationships accordingly.
If the spatial reference system (srid) of the point geometry is not the same as the topology, the apoint
is not a point geometry, the point is null, the point already exists as a node, the edge does not correspond to an existing edge or the point is not within the edge then an exception is thrown.
Availability: 1.?
This method implements the SQL/MM specification. SQL-MM: Topo-Net Routines: X.3.8
-- Add an edge -- SELECT topology.AddEdge('ma_topo', ST_GeomFromText('LINESTRING(227575 893917,227592 893900)', 26986) ) As edgeid; -- result- edgeid ------ 2 -- Split the new edge -- SELECT topology.ST_NewEdgesSplit('ma_topo', 2, ST_GeomFromText('POINT(227578.5 893913.5)', 26986) ) As newnodeid; newnodeid --------- 6
ST_RemoveIsoNode — Removes an isolated node and returns description of action. If the node is not isolated (is start or end of an edge), then an exception is thrown.
text ST_RemoveIsoNode(
varchar atopology, integer anode)
;
Removes an isolated node and returns description of action. If the node is not isolated (is start or end of an edge), then an exception is thrown.
Availability: 1.?
This method implements the SQL/MM specification. SQL-MM: Topo-Geo and Topo-Net 3: Routine Details: X+1.3.3
aface
.GetEdgeByPoint — Find the edge-id of an edge that intersects a given point
integer GetEdgeByPoint(
varchar atopology, geometry apoint, float8 tol)
;
The function returns an integer (id-edge) given a topology, a POINT and a tolerance. If tolerance = 0 then the point has to intersect the edge.
If the point doesn't intersect an edge, returns 0 (zero).
If use tolerance > 0 and there is more than one edge near the point then an exception is thrown.
If tolerance = 0, the function use ST_Intersects otherwise uses ST_DWithin. |
Availability: 2.0.0 - requires GEOS >= 3.3.0.
These examples use edges we created in AddEdge
SELECT topology.GetEdgeByPoint('ma_topo',geom, 1) As with1mtol, topology.GetEdgeByPoint('ma_topo',geom,0) As withnotol FROM ST_GeomFromEWKT('SRID=26986;POINT(227622.6 893843)') As geom; with1mtol | withnotol -----------+----------- 2 | 0
SELECT topology.GetEdgeByPoint('ma_topo',geom, 1) As nearnode FROM ST_GeomFromEWKT('SRID=26986;POINT(227591.9 893900.4)') As geom; -- get error -- ERROR: Two or more edges found
GetFaceByPoint — Find the face-id of a face that intersects a given point
integer GetFaceByPoint(
varchar atopology, geometry apoint, float8 tol)
;
Retrieve the id of a face that intersects a Point.
The function returns an integer (id-face) given a topology, a POINT and a tolerance. If tolerance = 0 then the point has to intersect the face.
If the point doesn't intersect a face, returns 0 (zero).
If use tolerance > 0 and there is more than one face near the point then an exception is thrown.
If tolerance = 0, the function uses ST_Intersects otherwise uses ST_DWithin. |
Availability: 2.0.0 - requires GEOS >= 3.3.0.
These examples use edges faces created in AddFace
SELECT topology.GetFaceByPoint('ma_topo',geom, 10) As with1mtol, topology.GetFaceByPoint('ma_topo',geom,0) As withnotol FROM ST_GeomFromEWKT('POINT(234604.6 899382.0)') As geom; with1mtol | withnotol -----------+----------- 1 | 0
SELECT topology.GetFaceByPoint('ma_topo',geom, 1) As nearnode FROM ST_GeomFromEWKT('POINT(227591.9 893900.4)') As geom; -- get error -- ERROR: Two or more faces found
GetNodeByPoint — Find the id of a node at a point location
integer GetNodeByPoint(
varchar atopology, geometry point, float8 tol)
;
The function return an integer (id-node) given a topology, a POINT and a tolerance. If tolerance = 0 mean exactly intersection otherwise retrieve the node from an interval.
If there isn't a node at the point, it return 0 (zero).
If use tolerance > 0 and near the point there are more than one node it throw an exception.
If tolerance = 0, the function use ST_Intersects otherwise will use ST_DWithin. |
Availability: 2.0.0 - requires GEOS >= 3.3.0.
These examples use edges we created in AddEdge
SELECT topology.GetNodeByPoint('ma_topo',geom, 1) As nearnode FROM ST_GeomFromEWKT('SRID=26986;POINT(227591.9 893900.4)') As geom; nearnode ---------- 2
SELECT topology.GetNodeByPoint('ma_topo',geom, 1000) As too_much_tolerance FROM ST_GeomFromEWKT('SRID=26986;POINT(227591.9 893900.4)') As geom; ----get error-- ERROR: Two or more nodes found
GetTopologyID — Returns the id of a topology in the topology.topology table given the name of the topology.
integer GetTopologyID(
varchar toponame)
;
GetTopologySRID — Returns the SRID of a topology in the topology.topology table given the name of the topology.
integer GetTopologyID(
varchar toponame)
;
GetTopologyName — Returns the name of a topology (schema) given the id of the topology.
varchar GetTopologyName(
integer topology_id)
;
ST_GetFaceEdges — Returns a set of ordered edges that bound aface
.
getfaceedges_returntype ST_GetFaceEdges(
varchar atopology, integer aface)
;
Returns a set of ordered edges that bound aface
. Each output consists of a sequence and edgeid. Sequence numbers start with value 1.
Enumeration of each ring edges start from the edge with smallest identifier. Order of edges follows a left-hand-rule (bound face is on the left of each directed edge).
Disponibilità: 2.0
This method implements the SQL/MM specification. SQL-MM 3 Topo-Geo and Topo-Net 3: Routine Details: X.3.5
-- Returns the edges bounding face 1 SELECT (topology.ST_GetFaceEdges('tt', 1)).*; -- result -- sequence | edge ----------+------ 1 | -4 2 | 5 3 | 7 4 | -6 5 | 1 6 | 2 7 | 3 (7 rows)
-- Returns the sequence, edge id -- and geometry of the edges that bound face 1 -- If you just need geom and seq, can use ST_GetFaceGeometry SELECT t.seq, t.edge, geom FROM topology.ST_GetFaceEdges('tt',1) As t(seq,edge) INNER JOIN tt.edge AS e ON abs(t.edge) = e.edge_id;
ST_GetFaceGeometry — Returns the polygon in the given topology with the specified face id.
geometry ST_GetFaceGeometry(
varchar atopology, integer aface)
;
Returns the polygon in the given topology with the specified face id. Builds the polygon from the edges making up the face.
Availability: 1.?
This method implements the SQL/MM specification. SQL-MM 3 Topo-Geo and Topo-Net 3: Routine Details: X.3.16
-- Returns the wkt of the polygon added with AddFace SELECT ST_AsText(topology.ST_GetFaceGeometry('ma_topo', 1)) As facegeomwkt; -- result -- facegeomwkt -------------------------------------------------------------------------------- POLYGON((234776.9 899563.7,234896.5 899456.7,234914 899436.4,234946.6 899356.9, 234872.5 899328.7,234891 899285.4,234992.5 899145,234890.6 899069, 234755.2 899255.4,234612.7 899379.4,234776.9 899563.7))
GetRingEdges — Returns the ordered set of signed edge identifiers met by walking on an a given edge side.
getfaceedges_returntype GetRingEdges(
varchar atopology, integer aring, integer max_edges=null)
;
Returns the ordered set of signed edge identifiers met by walking on an a given edge side. Each output consists of a sequence and a signed edge id. Sequence numbers start with value 1.
If you pass a positive edge id, the walk starts on the left side of the corresponding edge and follows the edge direction. If you pass a negative edge id, the walk starts on the right side of it and goes backward.
If max_edges
is not null no more than those records are returned by that function. This is meant to be a safety parameter when dealing with possibly invalid topologies.
This function uses edge ring linking metadata. |
Availability: 2.0.0
GetNodeEdges — Returns an ordered set of edges incident to the given node.
getfaceedges_returntype GetNodeEdges(
varchar atopology, integer anode)
;
Returns an ordered set of edges incident to the given node. Each output consists of a sequence and a signed edge id. Sequence numbers start with value 1. A positive edge starts at the given node. A negative edge ends into the given node. Closed edges will appear twice (with both signs). Order is clockwise starting from northbound.
This function computes ordering rather than deriving from metadata and is thus usable to build edge ring linking. |
Disponibilità: 2.0
Polygonize — Find and register all faces defined by topology edges
text Polygonize(
varchar toponame)
;
Register all faces that can be built out a topology edge primitives.
The target topology is assumed to contain no self-intersecting edges.
Already known faces are recognized, so it is safe to call Polygonize multiple times on the same topology. |
This function does not use nor set the next_left_edge and next_right_edge fields of the edge table. |
Availability: 2.0.0
AddNode — Adds a point node to the node table in the specified topology schema and returns the nodeid of new node. If point already exists as node, the existing nodeid is returned.
integer AddNode(
varchar toponame, geometry apoint, boolean allowEdgeSplitting=false, boolean computeContainingFace=false)
;
Adds a point node to the node table in the specified topology schema. The AddEdge function automatically adds start and end points of an edge when called so not necessary to explicitly add nodes of an edge.
If any edge crossing the node is found either an exception is raised or the edge is split, depending on the allowEdgeSplitting
parameter value.
If computeContainingFace
is true a newly added node would get the correct containing face computed.
If the |
Availability: 2.0.0
AddEdge — Adds a linestring edge to the edge table and associated start and end points to the point nodes table of the specified topology schema using the specified linestring geometry and returns the edgeid of the new (or existing) edge.
integer AddEdge(
varchar toponame, geometry aline)
;
Adds an edge to the edge table and associated nodes to the nodes table of the specified toponame
schema using the specified linestring geometry and returns the edgeid of the new or existing record. The newly added edge has "universe" face on both sides and links to itself.
If the |
The geometry of |
Availability: 2.0.0 requires GEOS >= 3.3.0.
SELECT topology.AddEdge('ma_topo', ST_GeomFromText('LINESTRING(227575.8 893917.2,227591.9 893900.4)', 26986) ) As edgeid; -- result- edgeid -------- 1 SELECT topology.AddEdge('ma_topo', ST_GeomFromText('LINESTRING(227591.9 893900.4,227622.6 893844.2,227641.6 893816.5, 227704.5 893778.5)', 26986) ) As edgeid; -- result -- edgeid -------- 2 SELECT topology.AddEdge('ma_topo', ST_GeomFromText('LINESTRING(227591.2 893900, 227591.9 893900.4, 227704.5 893778.5)', 26986) ) As edgeid; -- gives error -- ERROR: Edge intersects (not on endpoints) with existing edge 1
AddFace — Registers a face primitive to a topology and gets its identifier.
integer AddFace(
varchar toponame, geometry apolygon, boolean force_new=false)
;
Registers a face primitive to a topology and gets its identifier.
For a newly added face, the edges forming its boundaries and the ones contained in the face will be updated to have correct values in the left_face and right_face fields. Isolated nodes contained in the face will also be updated to have a correct containing_face field value.
This function does not use nor set the next_left_edge and next_right_edge fields of the edge table. |
The target topology is assumed to be valid (containing no self-intersecting edges). An exception is raised if: The polygon boundary is not fully defined by existing edges or the polygon overlaps an existing face.
If the apolygon
geometry already exists as a face, then: if force_new
is false (the default) the face id of the existing face is returned; if force_new
is true a new id will be assigned to the newly registered face.
When a new registration of an existing face is performed (force_new=true), no action will be taken to resolve dangling references to the existing face in the edge, node an relation tables, nor will the MBR field of the existing face record be updated. It is up to the caller to deal with that. |
The |
Availability: 2.0.0
-- first add the edges we use generate_series as an iterator (the below -- will only work for polygons with < 10000 points because of our max in gs) SELECT topology.AddEdge('ma_topo', ST_MakeLine(ST_PointN(geom,i), ST_PointN(geom, i + 1) )) As edgeid FROM (SELECT ST_NPoints(geom) AS npt, geom FROM (SELECT ST_Boundary(ST_GeomFromText('POLYGON((234896.5 899456.7,234914 899436.4,234946.6 899356.9,234872.5 899328.7, 234891 899285.4,234992.5 899145, 234890.6 899069,234755.2 899255.4, 234612.7 899379.4,234776.9 899563.7,234896.5 899456.7))', 26986) ) As geom ) As geoms) As facen CROSS JOIN generate_series(1,10000) As i WHERE i < npt; -- result -- edgeid -------- 3 4 5 6 7 8 9 10 11 12 (10 rows) -- then add the face - SELECT topology.AddFace('ma_topo', ST_GeomFromText('POLYGON((234896.5 899456.7,234914 899436.4,234946.6 899356.9,234872.5 899328.7, 234891 899285.4,234992.5 899145, 234890.6 899069,234755.2 899255.4, 234612.7 899379.4,234776.9 899563.7,234896.5 899456.7))', 26986) ) As faceid; -- result -- faceid -------- 1
ST_Simplify — Returns a "simplified" geometry version of the given TopoGeometry using the Douglas-Peucker algorithm.
geometry ST_Simplify(
TopoGeometry geomA, float tolerance)
;
Returns a "simplified" geometry version of the given TopoGeometry using the Douglas-Peucker algorithm on each component edge.
The returned geometry may be non-simple or non-valid. Splitting component edges may help retaining simplicity/validity. |
Performed by the GEOS module.
Availability: 2.1.0
topoelementarray
for a set of element_id, type arrays (topoelements)CreateTopoGeom — Creates a new topo geometry object from topo element array - tg_type: 1:[multi]point, 2:[multi]line, 3:[multi]poly, 4:collection
topogeometry CreateTopoGeom(
varchar toponame, integer tg_type, integer layer_id, topoelementarray tg_objs)
;
topogeometry CreateTopoGeom(
varchar toponame, integer tg_type, integer layer_id)
;
Creates a topogeometry object for layer denoted by layer_id and registers it in the relations table in the toponame
schema.
tg_type is an integer: 1:[multi]point (punctal), 2:[multi]line (lineal), 3:[multi]poly (areal), 4:collection. layer_id is the layer id in the topology.layer table.
punctal layers are formed from set of nodes, lineal layers are formed from a set of edges, areal layers are formed from a set of faces, and collections can be formed from a mixture of nodes, edges, and faces.
Omitting the array of components generates an empty TopoGeometry object.
Availability: 1.?
Create a topogeom in ri_topo schema for layer 2 (our ri_roads), of type (2) LINE, for the first edge (we loaded in ST_CreateTopoGeo
.
INSERT INTO ri.ri_roads(road_name, topo) VALUES('Unknown', topology.CreateTopoGeom('ri_topo',2,2,'{{1,2}}'::topology.topoelementarray);
Lets say we have geometries that should be formed from a collection of faces. We have for example blockgroups table and want to know the topo geometry of each block group. If our data was perfectly aligned, we could do this:
-- create our topo geometry column -- SELECT topology.AddTopoGeometryColumn( 'topo_boston', 'boston', 'blockgroups', 'topo', 'POLYGON'); -- addtopgeometrycolumn -- 1 -- update our column assuming -- everything is perfectly aligned with our edges UPDATE boston.blockgroups AS bg SET topo = topology.CreateTopoGeom('topo_boston' ,3,1 , foo.bfaces) FROM (SELECT b.gid, topology.TopoElementArray_Agg(ARRAY[f.face_id,3]) As bfaces FROM boston.blockgroups As b INNER JOIN topo_boston.face As f ON b.geom && f.mbr WHERE ST_Covers(b.geom, topology.ST_GetFaceGeometry('topo_boston', f.face_id)) GROUP BY b.gid) As foo WHERE foo.gid = bg.gid;
--the world is rarely perfect allow for some error --count the face if 50% of it falls -- within what we think is our blockgroup boundary UPDATE boston.blockgroups AS bg SET topo = topology.CreateTopoGeom('topo_boston' ,3,1 , foo.bfaces) FROM (SELECT b.gid, topology.TopoElementArray_Agg(ARRAY[f.face_id,3]) As bfaces FROM boston.blockgroups As b INNER JOIN topo_boston.face As f ON b.geom && f.mbr WHERE ST_Covers(b.geom, topology.ST_GetFaceGeometry('topo_boston', f.face_id)) OR ( ST_Intersects(b.geom, topology.ST_GetFaceGeometry('topo_boston', f.face_id)) AND ST_Area(ST_Intersection(b.geom, topology.ST_GetFaceGeometry('topo_boston', f.face_id) ) ) > ST_Area(topology.ST_GetFaceGeometry('topo_boston', f.face_id))*0.5 ) GROUP BY b.gid) As foo WHERE foo.gid = bg.gid; -- and if we wanted to convert our topogeometry back -- to a denomalized geometry aligned with our faces and edges -- cast the topo to a geometry -- The really cool thing is my new geometries -- are now aligned with my tiger street centerlines UPDATE boston.blockgroups SET new_geom = topo::geometry;
toTopoGeom — Converts a simple Geometry into a topo geometry
topogeometry toTopoGeom(
geometry geom, varchar toponame, integer layer_id, float8 tolerance)
;
topogeometry toTopoGeom(
geometry geom, topogeometry topogeom, float8 tolerance)
;
Converts a simple Geometry into a TopoGeometry.
Topological primitives required to represent the input geometry will be added to the underlying topology, possibly splitting existing ones, and they will be associated with the output TopoGeometry in the relation
table.
Existing TopoGeometry objects (with the possible exception of topogeom
, if given) will retain their shapes.
When tolerance
is given it will be used to snap the input geometry to existing primitives.
In the first form a new TopoGeometry will be created for the given layer (layer_id
) of the given topology (toponame
).
In the second form the primitives resulting from the conversion will be added to the pre-existing TopoGeometry (topogeom
), possibly adding space to its final shape. To have the new shape completely replace the old one see clearTopoGeom.
Disponibilità: 2.0
Enhanced: 2.1.0 adds the version taking an existing TopoGeometry.
This is a full self-contained workflow
-- do this if you don't have a topology setup already -- creates topology not allowing any tolerance SELECT topology.CreateTopology('topo_boston_test', 2249); -- create a new table CREATE TABLE nei_topo(gid serial primary key, nei varchar(30)); --add a topogeometry column to it SELECT topology.AddTopoGeometryColumn('topo_boston_test', 'public', 'nei_topo', 'topo', 'MULTIPOLYGON') As new_layer_id; new_layer_id ----------- 1 --use new layer id in populating the new topogeometry column -- we add the topogeoms to the new layer with 0 tolerance INSERT INTO nei_topo(nei, topo) SELECT nei, topology.toTopoGeom(geom, 'topo_boston_test', 1) FROM neighborhoods WHERE gid BETWEEN 1 and 15; --use to verify what has happened -- SELECT * FROM topology.TopologySummary('topo_boston_test'); -- summary-- Topology topo_boston_test (5), SRID 2249, precision 0 61 nodes, 87 edges, 35 faces, 15 topogeoms in 1 layers Layer 1, type Polygonal (3), 15 topogeoms Deploy: public.nei_topo.topo
-- Shrink all TopoGeometry polygons by 10 meters UPDATE nei_topo SET topo = ST_Buffer(clearTopoGeom(topo), -10); -- Get the no-one-lands left by the above operation -- I think GRASS calls this "polygon0 layer" SELECT ST_GetFaceGeometry('topo_boston_test', f.face_id) FROM topo_boston_test.face f WHERE f.face_id > 0 -- don't consider the universe face AND NOT EXISTS ( -- check that no TopoGeometry references the face SELECT * FROM topo_boston_test.relation WHERE layer_id = 1 AND element_id = f.face_id );
TopoElementArray_Agg — Returns a topoelementarray
for a set of element_id, type arrays (topoelements)
topoelementarray TopoElementArray_Agg(
topoelement set tefield)
;
clearTopoGeom — Clears the content of a topo geometry
topogeometry clearTopoGeom(
topogeometry topogeom)
;
Clears the content a TopoGeometry turning it into an empty one. Mostly useful in conjunction with toTopoGeom to replace the shape of existing objects and any dependent object in higher hierarchical levels.
Availability: 2.1
GetTopoGeomElements — Add an element to the definition of a TopoGeometry
geometry ST_GetFaceGeometry(
varchar atopology, integer aface)
;
Adds a TopoElement to the definition of a TopoGeometry object. Does not error out if the element is already part of the definition.
Disponibilità: 2.0
topoelementarray
(an array of topoelements) containing the topological elements and type of the given TopoGeometry (primitive elements)topoelement
objects containing the topological element_id,element_type of the given TopoGeometry (primitive elements)GetTopoGeomElementArray — Returns a topoelementarray
(an array of topoelements) containing the topological elements and type of the given TopoGeometry (primitive elements)
topoelementarray GetTopoGeomElementArray(
varchar toponame, integer layer_id, integer tg_id)
;
topoelementarray topoelement GetTopoGeomElementArray(
topogeometry tg)
;
Returns a TopoElementArray containing the topological elements and type of the given TopoGeometry (primitive elements). This is similar to GetTopoGeomElements except it returns the elements as an array rather than as a dataset.
tg_id is the topogeometry id of the topogeometry object in the topology in the layer denoted by layer_id
in the topology.layer table.
Availability: 1.?
GetTopoGeomElements — Returns a set of topoelement
objects containing the topological element_id,element_type of the given TopoGeometry (primitive elements)
setof topoelement GetTopoGeomElements(
varchar toponame, integer layer_id, integer tg_id)
;
setof topoelement GetTopoGeomElements(
topogeometry tg)
;
AsGML — Returns the GML representation of a topogeometry.
text AsGML(
topogeometry tg)
;
text AsGML(
topogeometry tg, text nsprefix_in)
;
text AsGML(
topogeometry tg, regclass visitedTable)
;
text AsGML(
topogeometry tg, regclass visitedTable, text nsprefix)
;
text AsGML(
topogeometry tg, text nsprefix_in, integer precision, integer options)
;
text AsGML(
topogeometry tg, text nsprefix_in, integer precision, integer options, regclass visitedTable)
;
text AsGML(
topogeometry tg, text nsprefix_in, integer precision, integer options, regclass visitedTable, text idprefix)
;
text AsGML(
topogeometry tg, text nsprefix_in, integer precision, integer options, regclass visitedTable, text idprefix, int gmlversion)
;
Returns the GML representation of a topogeometry in version GML3 format. If no nsprefix_in
is specified then gml
is used. Pass in an empty string for nsprefix to get a non-qualified name space. The precision (default: 15) and options (default 1) parameters, if given, are passed untouched to the underlying call to ST_AsGML.
The visitedTable
parameter, if given, is used for keeping track of the visited Node and Edge elements so to use cross-references (xlink:xref) rather than duplicating definitions. The table is expected to have (at least) two integer fields: 'element_type' and 'element_id'. The calling user must have both read and write privileges on the given table. For best performance, an index should be defined on element_type
and element_id
, in that order. Such index would be created automatically by adding a unique constraint to the fields. Example:
CREATE TABLE visited ( element_type integer, element_id integer, unique(element_type, element_id) );
The idprefix
parameter, if given, will be prepended to Edge and Node tag identifiers.
The gmlver
parameter, if given, will be passed to the underlying ST_AsGML. Defaults to 3.
Availability: 2.0.0
This uses the topo geometry we created in CreateTopoGeom
SELECT topology.AsGML(topo) As rdgml FROM ri.roads WHERE road_name = 'Unknown'; -- rdgml-- <gml:TopoCurve> <gml:directedEdge> <gml:Edge gml:id="E1"> <gml:directedNode orientation="-"> <gml:Node gml:id="N1"/> </gml:directedNode> <gml:directedNode ></gml:directedNode> <gml:curveProperty> <gml:Curve srsName="urn:ogc:def:crs:EPSG::3438"> <gml:segments> <gml:LineStringSegment> <gml:posList srsDimension="2" >384744 236928 384750 236923 384769 236911 384799 236895 384811 236890 384833 236884 384844 236882 384866 236881 384879 236883 384954 236898 385087 236932 385117 236938 385167 236938 385203 236941 385224 236946 385233 236950 385241 236956 385254 236971 385260 236979 385268 236999 385273 237018 385273 237037 385271 237047 385267 237057 385225 237125 385210 237144 385192 237161 385167 237192 385162 237202 385159 237214 385159 237227 385162 237241 385166 237256 385196 237324 385209 237345 385234 237375 385237 237383 385238 237399 385236 237407 385227 237419 385213 237430 385193 237439 385174 237451 385170 237455 385169 237460 385171 237475 385181 237503 385190 237521 385200 237533 385206 237538 385213 237541 385221 237542 385235 237540 385242 237541 385249 237544 385260 237555 385270 237570 385289 237584 385292 237589 385291 237596 385284 237630</gml:posList> </gml:LineStringSegment> </gml:segments> </gml:Curve> </gml:curveProperty> </gml:Edge> </gml:directedEdge> </gml:TopoCurve >
Same exercise as previous without namespace
SELECT topology.AsGML(topo,'') As rdgml FROM ri.roads WHERE road_name = 'Unknown'; -- rdgml-- <TopoCurve> <directedEdge> <Edge id="E1"> <directedNode orientation="-"> <Node id="N1"/> </directedNode> <directedNode ></directedNode> <curveProperty> <Curve srsName="urn:ogc:def:crs:EPSG::3438"> <segments> <LineStringSegment> <posList srsDimension="2" >384744 236928 384750 236923 384769 236911 384799 236895 384811 236890 384833 236884 384844 236882 384866 236881 384879 236883 384954 236898 385087 236932 385117 236938 385167 236938 385203 236941 385224 236946 385233 236950 385241 236956 385254 236971 385260 236979 385268 236999 385273 237018 385273 237037 385271 237047 385267 237057 385225 237125 385210 237144 385192 237161 385167 237192 385162 237202 385159 237214 385159 237227 385162 237241 385166 237256 385196 237324 385209 237345 385234 237375 385237 237383 385238 237399 385236 237407 385227 237419 385213 237430 385193 237439 385174 237451 385170 237455 385169 237460 385171 237475 385181 237503 385190 237521 385200 237533 385206 237538 385213 237541 385221 237542 385235 237540 385242 237541 385249 237544 385260 237555 385270 237570 385289 237584 385292 237589 385291 237596 385284 237630</posList> </LineStringSegment> </segments> </Curve> </curveProperty> </Edge> </directedEdge> </TopoCurve >
AsTopoJSON — Returns the TopoJSON representation of a topogeometry.
text AsTopoJSON(
topogeometry tg, regclass edgeMapTable)
;
Returns the TopoJSON representation of a topogeometry. If edgeMapTable
is not null, it will be used as a lookup/storage mapping of edge identifiers to arc indices. This is to be able to allow for a compact "arcs" array in the final document.
The table, if given, is expected to have an "arc_id" field of type "serial" and an "edge_id" of type integer; the code will query the table for "edge_id" so it is recommended to add an index on that field.
Arc indices in the TopoJSONjoutput are 0-based but they are 1-based in the "edgeMapTable" table. |
A full TopoJSON document will be need to contain, in addition to the snippets returned by this function, the actual arcs plus some headers. See the TopoJSON specification.
Availability: 2.1.0
Enhanced: 2.2.1 added support for puntal inputs
CREATE TEMP TABLE edgemap(arc_id serial, edge_id int unique); -- header SELECT '{ "type": "Topology", "transform": { "scale": [1,1], "translate": [0,0] }, "objects": {' -- objects UNION ALL SELECT '"' || feature_name || '": ' || AsTopoJSON(feature, 'edgemap') FROM features.big_parcels WHERE feature_name = 'P3P4'; -- arcs WITH edges AS ( SELECT m.arc_id, e.geom FROM edgemap m, city_data.edge e WHERE e.edge_id = m.edge_id ), points AS ( SELECT arc_id, (st_dumppoints(geom)).* FROM edges ), compare AS ( SELECT p2.arc_id, CASE WHEN p1.path IS NULL THEN p2.geom ELSE ST_Translate(p2.geom, -ST_X(p1.geom), -ST_Y(p1.geom)) END AS geom FROM points p2 LEFT OUTER JOIN points p1 ON ( p1.arc_id = p2.arc_id AND p2.path[1] = p1.path[1]+1 ) ORDER BY arc_id, p2.path ), arcsdump AS ( SELECT arc_id, (regexp_matches( ST_AsGeoJSON(geom), '\[.*\]'))[1] as t FROM compare ), arcs AS ( SELECT arc_id, '[' || array_to_string(array_agg(t), ',') || ']' as a FROM arcsdump GROUP BY arc_id ORDER BY arc_id ) SELECT '}, "arcs": [' UNION ALL SELECT array_to_string(array_agg(a), E',\n') from arcs -- footer UNION ALL SELECT ']}'::text as t; -- Result: { "type": "Topology", "transform": { "scale": [1,1], "translate": [0,0] }, "objects": { "P3P4": { "type": "MultiPolygon", "arcs": [[[-1]],[[6,5,-5,-4,-3,1]]]} }, "arcs": [ [[25,30],[6,0],[0,10],[-14,0],[0,-10],[8,0]], [[35,6],[0,8]], [[35,6],[12,0]], [[47,6],[0,8]], [[47,14],[0,8]], [[35,22],[12,0]], [[35,14],[0,8]] ]}
Equals — Returns true if two topogeometries are composed of the same topology primitives.
boolean Equals(
topogeometry tg1, topogeometry tg2)
;
Returns true if two topogeometries are composed of the same topology primitives: faces, edges, nodes.
This function not supported for topogeometries that are geometry collections. It also can not compare topogeometries from different topologies. |
Disponibilità: 1.1.0
This function supports 3d and will not drop the z-index.
Intersects — Returns true if any pair of primitives from the two topogeometries intersect.
boolean Intersects(
topogeometry tg1, topogeometry tg2)
;
Returns true if any pair of primitives from the two topogeometries intersect.
This function not supported for topogeometries that are geometry collections. It also can not compare topogeometries from different topologies. Also not currently supported for hierarchichal topogeometries (topogeometries composed of other topogeometries). |
Disponibilità: 1.1.0
This function supports 3d and will not drop the z-index.
Table of Contents
This is a fork of the PAGC standardizer (original code for this portion was PAGC PostgreSQL Address Standardizer).
The address standardizer is a single line address parser that takes an input address and normalizes it based on a set of rules stored in a table and helper lex and gaz tables.
The code is built into a single postgresql extension library called address_standardizer
which can be installed with CREATE EXTENSION address_standardizer;
. In addition to the address_standardizer extension, a sample data extension called address_standardizer_data_us
extensions is built, which contains gaz, lex, and rules tables for US data. This extensions can be installed via: CREATE EXTENSION address_standardizer_data_us;
The code for this extension can be found in the PostGIS extensions/address_standardizer
and is currently self-contained.
For installation instructions refer to: Section 2.7, “Installing and Using the address standardizer”.
The parser works from right to left looking first at the macro elements for postcode, state/province, city, and then looks micro elements to determine if we are dealing with a house number street or intersection or landmark. It currently does not look for a country code or name, but that could be introduced in the future.
Assumed to be US or CA based on: postcode as US or Canada state/province as US or Canada else US
These are recognized using Perl compatible regular expressions. These regexs are currently in the parseaddress-api.c and are relatively simple to make changes to if needed.
These are recognized using Perl compatible regular expressions. These regexs are currently in the parseaddress-api.c but could get moved into includes in the future for easier maintenance.
standardize_address
function.stdaddr — A composite type that consists of the elements of an address. This is the return type for standardize_address
function.
A composite type that consists of elements of an address. This is the return type for standardize_address function. Some descriptions for elements are borrowed from PAGC Postal Attributes.
The token numbers denote the output reference number in the rules table.
This method needs address_standardizer extension.
is text (token number 0
): Refers to building number or name. Unparsed building identifiers and types. Generally blank for most addresses.
is a text (token number 1
): This is the street number on a street. Example 75 in 75 State Street
.
is text (token number 2
): STREET NAME PRE-DIRECTIONAL such as North, South, East, West etc.
is text (token number 3
): STREET NAME PRE-MODIFIER Example OLD in 3715 OLD HIGHWAY 99
.
is text (token number 4
): STREET PREFIX TYPE
is text (token number 5
): STREET NAME
is text (token number 6
): STREET POST TYPE e.g. St, Ave, Cir. A street type following the root street name. Example STREET in 75 State Street
.
is text (token number 7
): STREET POST-DIRECTIONAL A directional modifier that follows the street name.. Example WEST in 3715 TENTH AVENUE WEST
.
is text (token number 8
): RURAL ROUTE . Example 8 in RR 7
.
is text: Extra information like Floor number.
is text (token number 10
): Example Boston.
is text (token number 11
): Example MASSACHUSETTS
is text (token number 12
): Example USA
is text POSTAL CODE (ZIP CODE) (token number 13
): Example 02109
is text POSTAL BOX NUMBER (token number 14 and 15
): Example 02109
is text Apartment number or Suite Number (token number 17
): Example 3B in APT 3B
.
Abstract
This section lists the PostgreSQL table formats used by the address_standardizer for normalizing addresses. Note that these tables do not need to be named the same as what is referenced here. You can have different lex, gaz, rules tables for each country for example or for your custom geocoder. The names of these tables get passed into the address standardizer functions.
The packaged extension address_standardizer_data_us
contains data for standardizing US addresses.
rules table — The rules table contains a set of rules that maps address input sequence tokens to standardized output sequence. A rule is defined as a set of input tokens followed by -1 (terminator) followed by set of output tokens followed by -1 followed by number denoting kind of rule followed by ranking of rule.
A rules table must have at least the following columns, though you are allowed to add more for your own uses.
Primary key of table
text field denoting the rule. Details at PAGC Address Standardizer Rule records.
A rule consists of a set of non-negative integers representing input tokens, terminated by a -1, followed by an equal number of non-negative integers representing postal attributes, terminated by a -1, followed by an integer representing a rule type, followed by an integer representing the rank of the rule. The rules are ranked from 0 (lowest) to 17 (highest).
So for example the rule 2 0 2 22 3 -1 5 5 6 7 3 -1 2 6
maps to sequence of output tokens TYPE NUMBER TYPE DIRECT QUALIF to the output sequence STREET STREET SUFTYP SUFDIR QUALIF. The rule is an ARC_C rule of rank 6.
Numbers for corresponding output tokens are listed in stdaddr.
Each rule starts with a set of input tokens followed by a terminator -1
. Valid input tokens excerpted from PAGC Input Tokens are as follows:
Form-Based Input Tokens
(13). The ampersand (&) is frequently used to abbreviate the word "and".
(9). A punctuation character.
(21). A sequence of two letters. Often used as identifiers.
(25). Fractions are sometimes used in civic numbers or unit numbers.
(23). An alphanumeric string that contains both letters and digits. Used for identifiers.
(0). A string of digits.
(15). Representations such as First or 1st. Often used in street names.
(18). A single letter.
(1). A word is a string of letters of arbitrary length. A single letter can be both a SINGLE and a WORD.
Function-based Input Tokens
(14). Words used to denote post office boxes. For example Box or PO Box.
(19). Words used to denote buildings or building complexes, usually as a prefix. For example: Tower in Tower 7A.
(24). Words and abbreviations used to denote buildings or building complexes, usually as a suffix. For example: Shopping Centre.
(22). Words used to denote directions, for example North.
(20). Words used to denote milepost addresses.
(6). Words and abbreviations used to denote highways and roads. For example: the Interstate in Interstate 5
(8). Words and abbreviations used to denote rural routes. RR.
(2). Words and abbreviation used to denote street typess. For example: ST or AVE.
(16). Words and abbreviation used to denote internal subaddresses. For example, APT or UNIT.
Postal Type Input Tokens
(28). A 5 digit number. Identifies a Zip Code
(29). A 4 digit number. Identifies ZIP4.
(27). A 3 character sequence of letter number letter. Identifies an FSA, the first 3 characters of a Canadian postal code.
(26). A 3 character sequence of number letter number. Identifies an LDU, the last 3 characters of a Canadian postal code.
Stopwords
STOPWORDS combine with WORDS. In rules a string of multiple WORDs and STOPWORDs will be represented by a single WORD token.
(7). A word with low lexical significance, that can be omitted in parsing. For example: THE.
After the first -1 (terminator), follows the output tokens and their order, followed by a terminator -1
. Numbers for corresponding output tokens are listed in stdaddr. What are allowed is dependent on kind of rule. Output tokens valid for each rule type are listed in the section called “Rule Types and Rank”.
The final part of the rule is the rule type which is denoted by one of the following, followed by a rule rank. The rules are ranked from 0 (lowest) to 17 (highest).
MACRO_C
(token number = "0"). The class of rules for parsing MACRO clauses such as PLACE STATE ZIP
MACRO_C output tokens (excerpted from http://www.pagcgeo.org/docs/html/pagc-12.html#--r-typ--.
(token number "10"). Example "Albany"
(token number "11"). Example "NY"
(token number "12"). This attribute is not used in most reference files. Example "USA"
(token number "13"). (SADS elements "ZIP CODE" , "PLUS 4" ). This attribute is used for both the US Zip and the Canadian Postal Codes.
MICRO_C
(token number = "1"). The class of rules for parsing full MICRO clauses (such as House, street, sufdir, predir, pretyp, suftype, qualif) (ie ARC_C plus CIVIC_C). These rules are not used in the build phase.
MICRO_C output tokens (excerpted from http://www.pagcgeo.org/docs/html/pagc-12.html#--r-typ--.
is a text (token number 1
): This is the street number on a street. Example 75 in 75 State Street
.
is text (token number 2
): STREET NAME PRE-DIRECTIONAL such as North, South, East, West etc.
is text (token number 3
): STREET NAME PRE-MODIFIER Example OLD in 3715 OLD HIGHWAY 99
.
is text (token number 4
): STREET PREFIX TYPE
is text (token number 5
): STREET NAME
is text (token number 6
): STREET POST TYPE e.g. St, Ave, Cir. A street type following the root street name. Example STREET in 75 State Street
.
is text (token number 7
): STREET POST-DIRECTIONAL A directional modifier that follows the street name.. Example WEST in 3715 TENTH AVENUE WEST
.
ARC_C
(token number = "2"). The class of rules for parsing MICRO clauses, excluding the HOUSE attribute. As such uses same set of output tokens as MICRO_C minus the HOUSE token.
CIVIC_C
(token number = "3"). The class of rules for parsing the HOUSE attribute.
EXTRA_C
(token number = "4"). The class of rules for parsing EXTRA attributes - attributes excluded from geocoding. These rules are not used in the build phase.
EXTRA_C output tokens (excerpted from http://www.pagcgeo.org/docs/html/pagc-12.html#--r-typ--.
(token number 0
): Unparsed building identifiers and types.
i(token number 14
): The BOX in BOX 3B
(token number 15
): The 3B in BOX 3B
(token number 8
): The RR in RR 7
(token number 16
): The APT in APT 3B
(token number 17
): The 3B in APT 3B
(token number 9
): An otherwise unclassified output.
lex table — A lex table is used to classify alphanumeric input and associate that input with (a) input tokens ( See the section called “Input Tokens”) and (b) standardized representations.
A lex (short for lexicon) table is used to classify alphanumeric input and associate that input with the section called “Input Tokens” and (b) standardized representations. Things you will find in these tables are ONE
mapped to stdword: 1
.
A lex has at least the following columns in the table. You may add
Primary key of table
integer: definition number?
text: the input word
text: the standardized replacement word
integer: the kind of word it is. Only if it is used in this context will it be replaced. Refer to PAGC Tokens.
gaz table — A gaz table is used to standardize place names and associate that input with (a) input tokens ( See the section called “Input Tokens”) and (b) standardized representations.
A gaz (short for gazeteer) table is used to classify place names and associate that input with the section called “Input Tokens” and (b) standardized representations. For example if you are in US, you may load these with State Names and associated abbreviations.
A gaz table has at least the following columns in the table. You may add more columns if you wish for your own purposes.
Primary key of table
integer: definition number? - identifer used for that instance of the word
text: the input word
text: the standardized replacement word
integer: the kind of word it is. Only if it is used in this context will it be replaced. Refer to PAGC Tokens.
parse_address — Takes a 1 line address and breaks into parts
record parse_address(
text address)
;
Returns takes as input an address, and returns a record output consisting of fields num, street, street2, address1, city, state, zip, zipplus, country.
Availability: 2.2.0
This method needs address_standardizer extension.
Single Addresss
SELECT num, street, city, zip, zipplus FROM parse_address('1 Devonshire Place, Boston, MA 02109-1234') AS a;
num | street | city | zip | zipplus -----+------------------+--------+-------+--------- 1 | Devonshire Place | Boston | 02109 | 1234
Table of addresses
-- basic table CREATE TABLE places(addid serial PRIMARY KEY, address text); INSERT INTO places(address) VALUES ('529 Main Street, Boston MA, 02129'), ('77 Massachusetts Avenue, Cambridge, MA 02139'), ('25 Wizard of Oz, Walaford, KS 99912323'), ('26 Capen Street, Medford, MA'), ('124 Mount Auburn St, Cambridge, Massachusetts 02138'), ('950 Main Street, Worcester, MA 01610'); -- parse the addresses -- if you want all fields you can use (a).* SELECT addid, (a).num, (a).street, (a).city, (a).state, (a).zip, (a).zipplus FROM (SELECT addid, parse_address(address) As a FROM places) AS p;
addid | num | street | city | state | zip | zipplus -------+-----+----------------------+-----------+-------+-------+--------- 1 | 529 | Main Street | Boston | MA | 02129 | 2 | 77 | Massachusetts Avenue | Cambridge | MA | 02139 | 3 | 25 | Wizard of Oz | Walaford | KS | 99912 | 323 4 | 26 | Capen Street | Medford | MA | | 5 | 124 | Mount Auburn St | Cambridge | MA | 02138 | 6 | 950 | Main Street | Worcester | MA | 01610 | (6 rows)
standardize_address — Returns an stdaddr form of an input address utilizing lex, gaz, and rule tables.
stdaddr standardize_address(
text lextab, text gaztab, text rultab, text address)
;
stdaddr standardize_address(
text lextab, text gaztab, text rultab, text micro, text macro)
;
Returns an stdaddr form of an input address utilizing lex table table name, gaz table, and rules table table names and an address.
Variant 1: Takes an address as a single line.
Variant 2: Takes an address as 2 parts. A micro
consisting of standard first line of postal address e.g. house_num street
, and a macro consisting of standard postal second line of an address e.g city, state postal_code country
.
Availability: 2.2.0
This method needs address_standardizer extension.
Using address_standardizer_data_us extension
CREATE EXTENSION address_standardizer_data_us; -- only needs to be done once
Variant 1: Single line address. This doesn't work well with non-US addresses
SELECT house_num, name, suftype, city, country, state, unit FROM standardize_address('us_lex', 'us_gaz', 'us_rules', 'One Devonshire Place, PH 301, Boston, MA 02109');
house_num | name | suftype | city | country | state | unit ----------+------------+---------+--------+---------+---------------+----------------- 1 | DEVONSHIRE | PLACE | BOSTON | USA | MASSACHUSETTS | # PENTHOUSE 301
Using tables packaged with tiger geocoder. This example only works if you installed postgis_tiger_geocoder
.
SELECT * FROM standardize_address('tiger.pagc_lex', 'tiger.pagc_gaz', 'tiger.pagc_rules', 'One Devonshire Place, PH 301, Boston, MA 02109-1234');
Make easier to read we'll dump output using hstore extension CREATE EXTENSION hstore; you need to install
SELECT (each(hstore(p))).* FROM standardize_address('tiger.pagc_lex', 'tiger.pagc_gaz', 'tiger.pagc_rules', 'One Devonshire Place, PH 301, Boston, MA 02109') As p;
key | value ------------+----------------- box | city | BOSTON name | DEVONSHIRE qual | unit | # PENTHOUSE 301 extra | state | MA predir | sufdir | country | USA pretype | suftype | PL building | postcode | 02109 house_num | 1 ruralroute | (16 rows)
Variant 2: As a two part Address
SELECT (each(hstore(p))).* FROM standardize_address('tiger.pagc_lex', 'tiger.pagc_gaz', 'tiger.pagc_rules', 'One Devonshire Place, PH 301', 'Boston, MA 02109, US') As p;
key | value ------------+----------------- box | city | BOSTON name | DEVONSHIRE qual | unit | # PENTHOUSE 301 extra | state | MA predir | sufdir | country | USA pretype | suftype | PL building | postcode | 02109 house_num | 1 ruralroute | (16 rows)
Table of Contents
This chapter documents features found in the extras folder of the PostGIS source tarballs and source repository. These are not always packaged with PostGIS binary releases, but are usually plpgsql based or standard shell scripts that can be run as is.
Abstract
A plpgsql based geocoder written to work with the TIGER (Topologically Integrated Geographic Encoding and Referencing system ) / Line and Master Address database export released by the US Census Bureau.
There are four components to the geocoder: the data loader functions, the address normalizer, the address geocoder, and the reverse geocoder.
Although it is designed specifically for the US, a lot of the concepts and functions are applicable and can be adapted to work with other country address and road networks.
The script builds a schema called tiger
to house all the tiger related functions, reusable lookup data such as road type prefixes, suffixes, states, various control tables for managing data load, and skeleton base tables from which all the tiger loaded tables inherit from.
Another schema called tiger_data
is also created which houses all the census data for each state that the loader downloads from Census site and loads into the database. In the current model, each set of state tables is prefixed with the state code e.g ma_addr
, ma_edges
etc with constraints to enforce only that state data. Each of these tables inherits from the tables addr
, faces
, edges
, etc located in the tiger schema
.
All the geocode functions only reference the base tables, so there is no requirement that the data schema be called tiger_data
or that data can't be further partitioned into other schemas -- e.g a different schema for each state, as long as all the tables inherit from the tables in the tiger
schema.
For instructions on how enable the extension in your database and also to load data using it, refer to Refer to Section 2.8.1, “Tiger Geocoder Enabling your PostGIS database: Using Extension”.
If you are using tiger geocoder (tiger_2010), you can upgrade the scripts using the accompanying upgrade_geocoder.bat / .sh scripts in extras/tiger. One major change between |
New in PostGIS 2.2.0 release is support for Tiger 2015 data and inclusion of Address Standardizer as part of PostGIS. New in PostGIS 2.1.0 release is ability to install tiger geocoder with PostgreSQL extension model if you are running PostgreSQL 9.1+. Refer to Section 2.8.1, “Tiger Geocoder Enabling your PostGIS database: Using Extension” for details. |
The Pagc_Normalize_Address function as a drop in replacement for in-built Normalize_Address. Refer to Section 2.7, “Installing and Using the address standardizer” for compile and installation instructions.
Design:
The goal of this project is to build a fully functional geocoder that can process an arbitrary United States address string and using normalized TIGER census data, produce a point geometry and rating reflecting the location of the given address and likeliness of the location. The higher the rating number the worse the result.
The reverse_geocode
function, introduced in PostGIS 2.0.0 is useful for deriving the street address and cross streets of a GPS location.
The geocoder should be simple for anyone familiar with PostGIS to install and use, and should be easily installable and usable on all platforms supported by PostGIS.
It should be robust enough to function properly despite formatting and spelling errors.
It should be extensible enough to be used with future data updates, or alternate data sources with a minimum of coding changes.
The |
tiger_data
if no schema is specified.county_all
, state_all
or stae code followed by county
or state
.tiger_data
if no schema is specified.tiger_data
schema. Each state script is returned as a separate record.tiger_data
schema. Each state script is returned as a separate record. Latest version supports Tiger 2010 structural changes and also loads census tract, block groups, and blocks tables.norm_addy
type that has road suffix, prefix and type standardized, street, streetname etc. broken into separate fields. This function will work with just the lookup data packaged with the tiger_geocoder (no need for tiger census data).norm_addy
type that has road suffix, prefix and type standardized, street, streetname etc. broken into separate fields. This function will work with just the lookup data packaged with the tiger_geocoder (no need for tiger census data). Requires address_standardizer extension.norm_addy
composite type object, returns a pretty print representation of it. Usually used in conjunction with normalize_address.There are a couple other open source geocoders for PostGIS, that unlike tiger geocoder have the advantage of multi-country geocoding support
Nominatim and uses OpenStreetMap gazeteer formatted data. It requires osm2pgsql for loading the data, PostgreSQL 8.4+ and PostGIS 1.5+ to function. It is packaged as a webservice interface and seems designed to be called as a webservice. Just like the tiger geocoder, it has both a geocoder and a reverse geocoder component. From the documentation, it is unclear if it has a pure SQL interface like the tiger geocoder, or if a good deal of the logic is implemented in the web interface.
GIS Graphy also utilizes PostGIS and like Nominatim works with OpenStreetMap (OSM) data. It comes with a loader to load OSM data and similar to Nominatim is capable of geocoding not just US. Much like Nominatim, it runs as a webservice and relies on Java 1.5, Servlet apps, Solr. GisGraphy is cross-platform and also has a reverse geocoder among some other neat features.
Drop_Indexes_Generate_Script — Generates a script that drops all non-primary key and non-unique indexes on tiger schema and user specified schema. Defaults schema to tiger_data
if no schema is specified.
text Drop_Indexes_Generate_Script(
text param_schema=tiger_data)
;
Generates a script that drops all non-primary key and non-unique indexes on tiger schema and user specified schema. Defaults schema to tiger_data
if no schema is specified.
This is useful for minimizing index bloat that may confuse the query planner or take up unnecessary space. Use in combination with Install_Missing_Indexes to add just the indexes used by the geocoder.
Availability: 2.0.0
SELECT drop_indexes_generate_script() As actionsql; actionsql --------------------------------------------------------- DROP INDEX tiger.idx_tiger_countysub_lookup_lower_name; DROP INDEX tiger.idx_tiger_edges_countyfp; DROP INDEX tiger.idx_tiger_faces_countyfp; DROP INDEX tiger.tiger_place_the_geom_gist; DROP INDEX tiger.tiger_edges_the_geom_gist; DROP INDEX tiger.tiger_state_the_geom_gist; DROP INDEX tiger.idx_tiger_addr_least_address; DROP INDEX tiger.idx_tiger_addr_tlid; DROP INDEX tiger.idx_tiger_addr_zip; DROP INDEX tiger.idx_tiger_county_countyfp; DROP INDEX tiger.idx_tiger_county_lookup_lower_name; DROP INDEX tiger.idx_tiger_county_lookup_snd_name; DROP INDEX tiger.idx_tiger_county_lower_name; DROP INDEX tiger.idx_tiger_county_snd_name; DROP INDEX tiger.idx_tiger_county_the_geom_gist; DROP INDEX tiger.idx_tiger_countysub_lookup_snd_name; DROP INDEX tiger.idx_tiger_cousub_countyfp; DROP INDEX tiger.idx_tiger_cousub_cousubfp; DROP INDEX tiger.idx_tiger_cousub_lower_name; DROP INDEX tiger.idx_tiger_cousub_snd_name; DROP INDEX tiger.idx_tiger_cousub_the_geom_gist; DROP INDEX tiger_data.idx_tiger_data_ma_addr_least_address; DROP INDEX tiger_data.idx_tiger_data_ma_addr_tlid; DROP INDEX tiger_data.idx_tiger_data_ma_addr_zip; DROP INDEX tiger_data.idx_tiger_data_ma_county_countyfp; DROP INDEX tiger_data.idx_tiger_data_ma_county_lookup_lower_name; DROP INDEX tiger_data.idx_tiger_data_ma_county_lookup_snd_name; DROP INDEX tiger_data.idx_tiger_data_ma_county_lower_name; DROP INDEX tiger_data.idx_tiger_data_ma_county_snd_name; : :
Drop_Nation_Tables_Generate_Script — Generates a script that drops all tables in the specified schema that start with county_all
, state_all
or stae code followed by county
or state
.
text Drop_Nation_Tables_Generate_Script(
text param_schema=tiger_data)
;
Generates a script that drops all tables in the specified schema that start with county_all
, state_all
or stae code followed by county
or state
. This is needed if you are upgrading from tiger_2010
to tiger_2011
data.
Availability: 2.1.0
Drop_State_Tables_Generate_Script — Generates a script that drops all tables in the specified schema that are prefixed with the state abbreviation. Defaults schema to tiger_data
if no schema is specified.
text Drop_State_Tables_Generate_Script(
text param_state, text param_schema=tiger_data)
;
Generates a script that drops all tables in the specified schema that are prefixed with the state abbreviation. Defaults schema to tiger_data
if no schema is specified. This function is useful for dropping tables of a state just before you reload a state in case something went wrong during your previous load.
Availability: 2.0.0
SELECT drop_state_tables_generate_script('PA'); DROP TABLE tiger_data.pa_addr; DROP TABLE tiger_data.pa_county; DROP TABLE tiger_data.pa_county_lookup; DROP TABLE tiger_data.pa_cousub; DROP TABLE tiger_data.pa_edges; DROP TABLE tiger_data.pa_faces; DROP TABLE tiger_data.pa_featnames; DROP TABLE tiger_data.pa_place; DROP TABLE tiger_data.pa_state; DROP TABLE tiger_data.pa_zip_lookup_base; DROP TABLE tiger_data.pa_zip_state; DROP TABLE tiger_data.pa_zip_state_loc;
Geocode — Takes in an address as a string (or other normalized address) and outputs a set of possible locations which include a point geometry in NAD 83 long lat, a normalized address for each, and the rating. The lower the rating the more likely the match. Results are sorted by lowest rating first. Can optionally pass in maximum results, defaults to 10, and restrict_region (defaults to NULL)
setof record geocode(
varchar address, integer max_results=10, geometry restrict_region=NULL, norm_addy OUT addy, geometry OUT geomout, integer OUT rating)
;
setof record geocode(
norm_addy in_addy, integer max_results=10, geometry restrict_region=NULL, norm_addy OUT addy, geometry OUT geomout, integer OUT rating)
;
Takes in an address as a string (or already normalized address) and outputs a set of possible locations which include a point geometry in NAD 83 long lat, a normalized_address
(addy) for each, and the rating. The lower the rating the more likely the match. Results are sorted by lowest rating first. Uses Tiger data (edges,faces,addr), PostgreSQL fuzzy string matching (soundex,levenshtein) and PostGIS line interpolation functions to interpolate address along the Tiger edges. The higher the rating the less likely the geocode is right. The geocoded point is defaulted to offset 10 meters from center-line off to side (L/R) of street address is located on.
Enhanced: 2.0.0 to support Tiger 2010 structured data and revised some logic to improve speed, accuracy of geocoding, and to offset point from centerline to side of street address is located on. New parameter max_results useful for specifying ot just return the best result.
The below examples timings are on a 3.0 GHZ single processor Windows 7 machine with 2GB ram running PostgreSQL 9.1rc1/PostGIS 2.0 loaded with all of MA,MN,CA, RI state Tiger data loaded.
Exact matches are faster to compute (61ms)
SELECT g.rating, ST_X(g.geomout) As lon, ST_Y(g.geomout) As lat, (addy).address As stno, (addy).streetname As street, (addy).streettypeabbrev As styp, (addy).location As city, (addy).stateabbrev As st,(addy).zip FROM geocode('75 State Street, Boston MA 02109') As g; rating | lon | lat | stno | street | styp | city | st | zip --------+-------------------+------------------+------+--------+------+--------+----+------- 0 | -71.0556722990239 | 42.3589914927049 | 75 | State | St | Boston | MA | 02109
Even if zip is not passed in the geocoder can guess (took about 122-150 ms)
SELECT g.rating, ST_AsText(ST_SnapToGrid(g.geomout,0.00001)) As wktlonlat, (addy).address As stno, (addy).streetname As street, (addy).streettypeabbrev As styp, (addy).location As city, (addy).stateabbrev As st,(addy).zip FROM geocode('226 Hanover Street, Boston, MA',1) As g; rating | wktlonlat | stno | street | styp | city | st | zip --------+---------------------------+------+---------+------+--------+----+------- 1 | POINT(-71.05528 42.36316) | 226 | Hanover | St | Boston | MA | 02113
Can handle misspellings and provides more than one possible solution with ratings and takes longer (500ms).
SELECT g.rating, ST_AsText(ST_SnapToGrid(g.geomout,0.00001)) As wktlonlat, (addy).address As stno, (addy).streetname As street, (addy).streettypeabbrev As styp, (addy).location As city, (addy).stateabbrev As st,(addy).zip FROM geocode('31 - 37 Stewart Street, Boston, MA 02116') As g; rating | wktlonlat | stno | street | styp | city | st | zip --------+---------------------------+------+--------+------+--------+----+------- 70 | POINT(-71.06459 42.35113) | 31 | Stuart | St | Boston | MA | 02116
Using to do a batch geocode of addresses. Easiest is to set max_results=1
. Only process those not yet geocoded (have no rating).
CREATE TABLE addresses_to_geocode(addid serial PRIMARY KEY, address text, lon numeric, lat numeric, new_address text, rating integer); INSERT INTO addresses_to_geocode(address) VALUES ('529 Main Street, Boston MA, 02129'), ('77 Massachusetts Avenue, Cambridge, MA 02139'), ('25 Wizard of Oz, Walaford, KS 99912323'), ('26 Capen Street, Medford, MA'), ('124 Mount Auburn St, Cambridge, Massachusetts 02138'), ('950 Main Street, Worcester, MA 01610'); -- only update the first 3 addresses (323-704 ms - there are caching and shared memory effects so first geocode you do is always slower) -- -- for large numbers of addresses you don't want to update all at once -- since the whole geocode must commit at once -- For this example we rejoin with LEFT JOIN -- and set to rating to -1 rating if no match -- to ensure we don't regeocode a bad address UPDATE addresses_to_geocode SET (rating, new_address, lon, lat) = ( COALESCE((g.geo).rating,-1), pprint_addy((g.geo).addy), ST_X((g.geo).geomout)::numeric(8,5), ST_Y((g.geo).geomout)::numeric(8,5) ) FROM (SELECT addid FROM addresses_to_geocode WHERE rating IS NULL ORDER BY addid LIMIT 3) As a LEFT JOIN (SELECT addid, (geocode(address,1)) As geo FROM addresses_to_geocode As ag WHERE ag.rating IS NULL ORDER BY addid LIMIT 3) As g ON a.addid = g.addid WHERE a.addid = addresses_to_geocode.addid; result ----- Query returned successfully: 3 rows affected, 480 ms execution time. SELECT * FROM addresses_to_geocode WHERE rating is not null; addid | address | lon | lat | new_address | rating -------+----------------------------------------------+-----------+----------+-------------------------------------------+-------- 1 | 529 Main Street, Boston MA, 02129 | -71.07181 | 42.38359 | 529 Main St, Boston, MA 02129 | 0 2 | 77 Massachusetts Avenue, Cambridge, MA 02139 | -71.09428 | 42.35988 | 77 Massachusetts Ave, Cambridge, MA 02139 | 0 3 | 25 Wizard of Oz, Walaford, KS 99912323 | | | | -1
SELECT g.rating, ST_AsText(ST_SnapToGrid(g.geomout,0.00001)) As wktlonlat, (addy).address As stno, (addy).streetname As street, (addy).streettypeabbrev As styp, (addy).location As city, (addy).stateabbrev As st,(addy).zip FROM geocode('100 Federal Street, MA', 3, (SELECT ST_Union(the_geom) FROM place WHERE statefp = '25' AND name = 'Lynn')::geometry ) As g; rating | wktlonlat | stno | street | styp | city | st | zip --------+--------------------------+------+---------+------+------+----+------- 8 | POINT(-70.96796 42.4659) | 100 | Federal | St | Lynn | MA | 01905 Total query runtime: 245 ms.
Geocode_Intersection — Takes in 2 streets that intersect and a state, city, zip, and outputs a set of possible locations on the first cross street that is at the intersection, also includes a point geometry in NAD 83 long lat, a normalized address for each location, and the rating. The lower the rating the more likely the match. Results are sorted by lowest rating first. Can optionally pass in maximum results, defaults to 10
setof record geocode_intersection(
text roadway1, text roadway2, text in_state, text in_city, text in_zip, integer max_results=10, norm_addy OUT addy, geometry OUT geomout, integer OUT rating)
;
Takes in 2 streets that intersect and a state, city, zip, and outputs a set of possible locations on the first cross street that is at the intersection, also includes a point geometry in NAD 83 long lat, a normalized address for each location, and the rating. The lower the rating the more likely the match. Results are sorted by lowest rating first. Can optionally pass in maximum results, defaults to 10. Returns normalized_address
(addy) for each, geomout as the point location in nad 83 long lat, and the rating. The lower the rating the more likely the match. Results are sorted by lowest rating first. Uses Tiger data (edges,faces,addr), PostgreSQL fuzzy string matching (soundex,levenshtein)
Availability: 2.0.0
The below examples timings are on a 3.0 GHZ single processor Windows 7 machine with 2GB ram running PostgreSQL 9.0/PostGIS 1.5 loaded with all of MA state Tiger data loaded. Currently a bit slow (3000 ms)
Testing on Windows 2003 64-bit 8GB on PostGIS 2.0 PostgreSQL 64-bit Tiger 2011 data loaded -- (41ms)
SELECT pprint_addy(addy), st_astext(geomout),rating FROM geocode_intersection( 'Haverford St','Germania St', 'MA', 'Boston', '02130',1); pprint_addy | st_astext | rating ----------------------------------+----------------------------+-------- 98 Haverford St, Boston, MA 02130 | POINT(-71.101375 42.31376) | 0
Even if zip is not passed in the geocoder can guess (took about 3500 ms on the windows 7 box), on the windows 2003 64-bit 741 ms
SELECT pprint_addy(addy), st_astext(geomout),rating FROM geocode_intersection('Weld', 'School', 'MA', 'Boston'); pprint_addy | st_astext | rating -------------------------------+--------------------------+-------- 98 Weld Ave, Boston, MA 02119 | POINT(-71.099 42.314234) | 3 99 Weld Ave, Boston, MA 02119 | POINT(-71.099 42.314234) | 3
Get_Geocode_Setting — Returns value of specific setting stored in tiger.geocode_settings table.
text Get_Geocode_Setting(
text setting_name)
;
Returns value of specific setting stored in tiger.geocode_settings table. Settings allow you to toggle debugging of functions. Later plans will be to control rating with settings. Current list of settings are as follows:
name | setting | unit | category | short_desc --------------------------------+---------+---------+-----------+------------------------------------------------------------------------------------------------------------------------------ debug_geocode_address | false | boolean | debug | outputs debug information in notice log such as queries when geocode_address is called if true debug_geocode_intersection | false | boolean | debug | outputs debug information in notice log such as queries when geocode_intersection is called if true debug_normalize_address | false | boolean | debug | outputs debug information in notice log such as queries and intermediate expressions when normalize_address is called if true debug_reverse_geocode | false | boolean | debug | if true, outputs debug information in notice log such as queries and intermediate expressions when reverse_geocode reverse_geocode_numbered_roads | 0 | integer | rating | For state and county highways, 0 - no preference in name, 1 - prefer the numbered highway name, 2 - prefer local state/county name use_pagc_address_parser | false | boolean | normalize | If set to true, will try to use the address_standardizer extension (via pagc_normalize_address) instead of tiger normalize_address built one
Changed: 2.2.0 : default settings are now kept in a table called geocode_settings_default. Use customized settingsa re in geocode_settings and only contain those that have been set by user.
Availability: 2.1.0
Get_Tract — Returns census tract or field from tract table of where the geometry is located. Default to returning short name of tract.
text get_tract(
geometry loc_geom, text output_field=name)
;
Given a geometry will return the census tract location of that geometry. NAD 83 long lat is assumed if no spatial ref sys is specified.
Availability: 2.0.0
Install_Missing_Indexes — Finds all tables with key columns used in geocoder joins and filter conditions that are missing used indexes on those columns and will add them.
boolean Install_Missing_Indexes(
)
;
Finds all tables in tiger
and tiger_data
schemas with key columns used in geocoder joins and filters that are missing indexes on those columns and will output the SQL DDL to define the index for those tables and then execute the generated script. This is a helper function that adds new indexes needed to make queries faster that may have been missing during the load process. This function is a companion to Missing_Indexes_Generate_Script that in addition to generating the create index script, also executes it. It is called as part of the update_geocode.sql
upgrade script.
Availability: 2.0.0
Loader_Generate_Census_Script — Generates a shell script for the specified platform for the specified states that will download Tiger census state tract, bg, and tabblocks data tables, stage and load into tiger_data
schema. Each state script is returned as a separate record.
setof text loader_generate_census_script(
text[] param_states, text os)
;
Generates a shell script for the specified platform for the specified states that will download Tiger data census state tract
, block groups bg
, and tabblocks
data tables, stage and load into tiger_data
schema. Each state script is returned as a separate record.
It uses unzip on Linux (7-zip on Windows by default) and wget to do the downloading. It uses Section 4.4.2, “shp2pgsql: Using the ESRI Shapefile Loader” to load in the data. Note the smallest unit it does is a whole state. It will only process the files in the staging and temp folders.
It uses the following control tables to control the process and different OS shell syntax variations.
loader_variables
keeps track of various variables such as census site, year, data and staging schemas
loader_platform
profiles of various platforms and where the various executables are located. Comes with windows and linux. More can be added.
loader_lookuptables
each record defines a kind of table (state, county), whether to process records in it and how to load them in. Defines the steps to import data, stage data, add, removes columns, indexes, and constraints for each. Each table is prefixed with the state and inherits from a table in the tiger schema. e.g. creates tiger_data.ma_faces
which inherits from tiger.faces
Availability: 2.0.0
Loader_Generate_Script includes this logic, but if you installed tiger geocoder prior to PostGIS 2.0.0 alpha5, you'll need to run this on the states you have already done to get these additional tables. |
Generate script to load up data for select states in Windows shell script format.
SELECT loader_generate_census_script(ARRAY['MA'], 'windows'); -- result -- set STATEDIR="\gisdata\www2.census.gov\geo\pvs\tiger2010st\25_Massachusetts" set TMPDIR=\gisdata\temp\ set UNZIPTOOL="C:\Program Files\7-Zip\7z.exe" set WGETTOOL="C:\wget\wget.exe" set PGBIN=C:\projects\pg\pg91win\bin\ set PGPORT=5432 set PGHOST=localhost set PGUSER=postgres set PGPASSWORD=yourpasswordhere set PGDATABASE=tiger_postgis20 set PSQL="%PGBIN%psql" set SHP2PGSQL="%PGBIN%shp2pgsql" cd \gisdata %WGETTOOL% http://www2.census.gov/geo/pvs/tiger2010st/25_Massachusetts/25/ --no-parent --relative --accept=*bg10.zip,*tract10.zip,*tabblock10.zip --mirror --reject=html del %TMPDIR%\*.* /Q %PSQL% -c "DROP SCHEMA tiger_staging CASCADE;" %PSQL% -c "CREATE SCHEMA tiger_staging;" cd %STATEDIR% for /r %%z in (*.zip) do %UNZIPTOOL% e %%z -o%TMPDIR% cd %TMPDIR% %PSQL% -c "CREATE TABLE tiger_data.MA_tract(CONSTRAINT pk_MA_tract PRIMARY KEY (tract_id) ) INHERITS(tiger.tract); " %SHP2PGSQL% -c -s 4269 -g the_geom -W "latin1" tl_2010_25_tract10.dbf tiger_staging.ma_tract10 | %PSQL% %PSQL% -c "ALTER TABLE tiger_staging.MA_tract10 RENAME geoid10 TO tract_id; SELECT loader_load_staged_data(lower('MA_tract10'), lower('MA_tract')); " %PSQL% -c "CREATE INDEX tiger_data_MA_tract_the_geom_gist ON tiger_data.MA_tract USING gist(the_geom);" %PSQL% -c "VACUUM ANALYZE tiger_data.MA_tract;" %PSQL% -c "ALTER TABLE tiger_data.MA_tract ADD CONSTRAINT chk_statefp CHECK (statefp = '25');" :
Generate sh script
STATEDIR="/gisdata/www2.census.gov/geo/pvs/tiger2010st/25_Massachusetts" TMPDIR="/gisdata/temp/" UNZIPTOOL=unzip WGETTOOL="/usr/bin/wget" export PGBIN=/usr/pgsql-9.0/bin export PGPORT=5432 export PGHOST=localhost export PGUSER=postgres export PGPASSWORD=yourpasswordhere export PGDATABASE=geocoder PSQL=${PGBIN}/psql SHP2PGSQL=${PGBIN}/shp2pgsql cd /gisdata wget http://www2.census.gov/geo/pvs/tiger2010st/25_Massachusetts/25/ --no-parent --relative --accept=*bg10.zip,*tract10.zip,*tabblock10.zip --mirror --reject=html rm -f ${TMPDIR}/*.* ${PSQL} -c "DROP SCHEMA tiger_staging CASCADE;" ${PSQL} -c "CREATE SCHEMA tiger_staging;" cd $STATEDIR for z in *.zip; do $UNZIPTOOL -o -d $TMPDIR $z; done : :
Loader_Generate_Script — Generates a shell script for the specified platform for the specified states that will download Tiger data, stage and load into tiger_data
schema. Each state script is returned as a separate record. Latest version supports Tiger 2010 structural changes and also loads census tract, block groups, and blocks tables.
setof text loader_generate_script(
text[] param_states, text os)
;
Generates a shell script for the specified platform for the specified states that will download Tiger data, stage and load into tiger_data
schema. Each state script is returned as a separate record.
It uses unzip on Linux (7-zip on Windows by default) and wget to do the downloading. It uses Section 4.4.2, “shp2pgsql: Using the ESRI Shapefile Loader” to load in the data. Note the smallest unit it does is a whole state, but you can overwrite this by downloading the files yourself. It will only process the files in the staging and temp folders.
It uses the following control tables to control the process and different OS shell syntax variations.
loader_variables
keeps track of various variables such as census site, year, data and staging schemas
loader_platform
profiles of various platforms and where the various executables are located. Comes with windows and linux. More can be added.
loader_lookuptables
each record defines a kind of table (state, county), whether to process records in it and how to load them in. Defines the steps to import data, stage data, add, removes columns, indexes, and constraints for each. Each table is prefixed with the state and inherits from a table in the tiger schema. e.g. creates tiger_data.ma_faces
which inherits from tiger.faces
Availability: 2.0.0 to support Tiger 2010 structured data and load census tract (tract), block groups (bg), and blocks (tabblocks) tables .
Generate script to load up data for 2 states in Windows shell script format.
SELECT loader_generate_script(ARRAY['MA','RI'], 'windows') AS result; -- result -- set STATEDIR="\gisdata\www2.census.gov\geo\pvs\tiger2010st\44_Rhode_Island" set TMPDIR=\gisdata\temp\ set UNZIPTOOL="C:\Program Files\7-Zip\7z.exe" set WGETTOOL="C:\wget\wget.exe" set PGBIN=C:\Program Files\PostgreSQL\8.4\bin\ set PGPORT=5432 set PGHOST=localhost set PGUSER=postgres set PGPASSWORD=yourpasswordhere set PGDATABASE=geocoder set PSQL="%PGBIN%psql" set SHP2PGSQL="%PGBIN%shp2pgsql" %WGETTOOL% http://www2.census.gov/geo/pvs/tiger2010st/44_Rhode_Island/ --no-parent --relative --recursive --level=2 --accept=zip,txt --mirror --reject=html : :
Generate sh script
SELECT loader_generate_script(ARRAY['MA','RI'], 'sh') AS result; -- result -- STATEDIR="/gisdata/www2.census.gov/geo/pvs/tiger2010st/44_Rhode_Island" TMPDIR="/gisdata/temp/" UNZIPTOOL=unzip PGPORT=5432 PGHOST=localhost PGUSER=postgres PGPASSWORD=yourpasswordhere PGDATABASE=geocoder PSQL=psql SHP2PGSQ=shp2pgsql wget http://www2.census.gov/geo/pvs/tiger2010st/44_Rhode_Island/ --no-parent --relative --recursive --level=2 --accept=zip,txt --mirror --reject=html : :
Loader_Generate_Nation_Script — Generates a shell script for the specified platform that loads in the county and state lookup tables.
text loader_generate_nation_script(
text os)
;
Generates a shell script for the specified platform that loads in the county_all
, county_all_lookup
, state_all
tables into tiger_data
schema. These inherit respectively from the county
, county_lookup
, state
tables in tiger
schema.
It uses unzip on Linux (7-zip on Windows by default) and wget to do the downloading. It uses Section 4.4.2, “shp2pgsql: Using the ESRI Shapefile Loader” to load in the data.
It uses the following control tables tiger.loader_platform
, tiger.loader_variables
, and tiger.loader_lookuptables
to control the process and different OS shell syntax variations.
loader_variables
keeps track of various variables such as census site, year, data and staging schemas
loader_platform
profiles of various platforms and where the various executables are located. Comes with windows and linux/unix. More can be added.
loader_lookuptables
each record defines a kind of table (state, county), whether to process records in it and how to load them in. Defines the steps to import data, stage data, add, removes columns, indexes, and constraints for each. Each table is prefixed with the state and inherits from a table in the tiger schema. e.g. creates tiger_data.ma_faces
which inherits from tiger.faces
Availability: 2.1.0
If you were running |
Missing_Indexes_Generate_Script — Finds all tables with key columns used in geocoder joins that are missing indexes on those columns and will output the SQL DDL to define the index for those tables.
text Missing_Indexes_Generate_Script(
)
;
Finds all tables in tiger
and tiger_data
schemas with key columns used in geocoder joins that are missing indexes on those columns and will output the SQL DDL to define the index for those tables. This is a helper function that adds new indexes needed to make queries faster that may have been missing during the load process. As the geocoder is improved, this function will be updated to accommodate new indexes being used. If this function outputs nothing, it means all your tables have what we think are the key indexes already in place.
Availability: 2.0.0
SELECT missing_indexes_generate_script(); -- output: This was run on a database that was created before many corrections were made to the loading script --- CREATE INDEX idx_tiger_county_countyfp ON tiger.county USING btree(countyfp); CREATE INDEX idx_tiger_cousub_countyfp ON tiger.cousub USING btree(countyfp); CREATE INDEX idx_tiger_edges_tfidr ON tiger.edges USING btree(tfidr); CREATE INDEX idx_tiger_edges_tfidl ON tiger.edges USING btree(tfidl); CREATE INDEX idx_tiger_zip_lookup_all_zip ON tiger.zip_lookup_all USING btree(zip); CREATE INDEX idx_tiger_data_ma_county_countyfp ON tiger_data.ma_county USING btree(countyfp); CREATE INDEX idx_tiger_data_ma_cousub_countyfp ON tiger_data.ma_cousub USING btree(countyfp); CREATE INDEX idx_tiger_data_ma_edges_countyfp ON tiger_data.ma_edges USING btree(countyfp); CREATE INDEX idx_tiger_data_ma_faces_countyfp ON tiger_data.ma_faces USING btree(countyfp);
Normalize_Address — Given a textual street address, returns a composite norm_addy
type that has road suffix, prefix and type standardized, street, streetname etc. broken into separate fields. This function will work with just the lookup data packaged with the tiger_geocoder (no need for tiger census data).
norm_addy normalize_address(
varchar in_address)
;
Given a textual street address, returns a composite norm_addy
type that has road suffix, prefix and type standardized, street, streetname etc. broken into separate fields. This is the first step in the geocoding process to get all addresses into normalized postal form. No other data is required aside from what is packaged with the geocoder.
This function just uses the various direction/state/suffix lookup tables preloaded with the tiger_geocoder and located in the tiger
schema, so it doesn't need you to download tiger census data or any other additional data to make use of it. You may find the need to add more abbreviations or alternative namings to the various lookup tables in the tiger
schema.
It uses various control lookup tables located in tiger
schema to normalize the input address.
Fields in the norm_addy
type object returned by this function in this order where () indicates a field required by the geocoder, [] indicates an optional field:
(address) [predirAbbrev] (streetName) [streetTypeAbbrev] [postdirAbbrev] [internal] [location] [stateAbbrev] [zip]
address
is an integer: The street number
predirAbbrev
is varchar: Directional prefix of road such as N, S, E, W etc. These are controlled using the direction_lookup
table.
streetName
varchar
streetTypeAbbrev
varchar abbreviated version of street type: e.g. St, Ave, Cir. These are controlled using the street_type_lookup
table.
postdirAbbrev
varchar abbreviated directional suffice of road N, S, E, W etc. These are controlled using the direction_lookup
table.
internal
varchar internal address such as an apartment or suite number.
location
varchar usually a city or governing province.
stateAbbrev
varchar two character US State. e.g MA, NY, MI. These are controlled by the state_lookup
table.
zip
varchar 5-digit zipcode. e.g. 02109.
parsed
boolean - denotes if addess was formed from normalize process. The normalize_address function sets this to true before returning the address.
Output select fields. Use Pprint_Addy if you want a pretty textual output.
SELECT address As orig, (g.na).streetname, (g.na).streettypeabbrev FROM (SELECT address, normalize_address(address) As na FROM addresses_to_geocode) As g; orig | streetname | streettypeabbrev -----------------------------------------------------+---------------+------------------ 28 Capen Street, Medford, MA | Capen | St 124 Mount Auburn St, Cambridge, Massachusetts 02138 | Mount Auburn | St 950 Main Street, Worcester, MA 01610 | Main | St 529 Main Street, Boston MA, 02129 | Main | St 77 Massachusetts Avenue, Cambridge, MA 02139 | Massachusetts | Ave 25 Wizard of Oz, Walaford, KS 99912323 | Wizard of Oz |
Pagc_Normalize_Address — Given a textual street address, returns a composite norm_addy
type that has road suffix, prefix and type standardized, street, streetname etc. broken into separate fields. This function will work with just the lookup data packaged with the tiger_geocoder (no need for tiger census data). Requires address_standardizer extension.
norm_addy pagc_normalize_address(
varchar in_address)
;
Given a textual street address, returns a composite norm_addy
type that has road suffix, prefix and type standardized, street, streetname etc. broken into separate fields. This is the first step in the geocoding process to get all addresses into normalized postal form. No other data is required aside from what is packaged with the geocoder.
This function just uses the various pagc_* lookup tables preloaded with the tiger_geocoder and located in the tiger
schema, so it doesn't need you to download tiger census data or any other additional data to make use of it. You may find the need to add more abbreviations or alternative namings to the various lookup tables in the tiger
schema.
It uses various control lookup tables located in tiger
schema to normalize the input address.
Fields in the norm_addy
type object returned by this function in this order where () indicates a field required by the geocoder, [] indicates an optional field:
There are slight variations in casing and formatting over the Normalize_Address.
Availability: 2.1.0
This method needs address_standardizer extension.
(address) [predirAbbrev] (streetName) [streetTypeAbbrev] [postdirAbbrev] [internal] [location] [stateAbbrev] [zip]
The native standardaddr of address_standardizer extension is at this time a bit richer than norm_addy since its designed to support international addresses (including country). standardaddr equivalent fields are:
house_num,predir, name, suftype, sufdir, unit, city, state, postcode
address
is an integer: The street number
predirAbbrev
is varchar: Directional prefix of road such as N, S, E, W etc. These are controlled using the direction_lookup
table.
streetName
varchar
streetTypeAbbrev
varchar abbreviated version of street type: e.g. St, Ave, Cir. These are controlled using the street_type_lookup
table.
postdirAbbrev
varchar abbreviated directional suffice of road N, S, E, W etc. These are controlled using the direction_lookup
table.
internal
varchar internal address such as an apartment or suite number.
location
varchar usually a city or governing province.
stateAbbrev
varchar two character US State. e.g MA, NY, MI. These are controlled by the state_lookup
table.
zip
varchar 5-digit zipcode. e.g. 02109.
parsed
boolean - denotes if addess was formed from normalize process. The normalize_address function sets this to true before returning the address.
Single call example
SELECT addy.* FROM pagc_normalize_address('9000 E ROO ST STE 999, Springfield, CO') AS addy; address | predirabbrev | streetname | streettypeabbrev | postdirabbrev | internal | location | stateabbrev | zip | parsed ---------+--------------+------------+------------------+---------------+-----------+-------------+-------------+-----+-------- 9000 | E | ROO | ST | | SUITE 999 | SPRINGFIELD | CO | | t
Batch call. There are currently speed issues with the way postgis_tiger_geocoder wraps the address_standardizer. These will hopefully be resolved in later editions. To work around them, if you need speed for batch geocoding to call generate a normaddy in batch mode, you are encouraged to directly call the address_standardizer standardize_address function as shown below which is similar exercise to what we did in Normalize_Address that uses data created in Geocode.
WITH g AS (SELECT address, ROW((sa).house_num, (sa).predir, (sa).name , (sa).suftype, (sa).sufdir, (sa).unit , (sa).city, (sa).state, (sa).postcode, true)::norm_addy As na FROM (SELECT address, standardize_address('tiger.pagc_lex' , 'tiger.pagc_gaz' , 'tiger.pagc_rules', address) As sa FROM addresses_to_geocode) As g) SELECT address As orig, (g.na).streetname, (g.na).streettypeabbrev FROM g; orig | streetname | streettypeabbrev -----------------------------------------------------+---------------+------------------ 529 Main Street, Boston MA, 02129 | MAIN | ST 77 Massachusetts Avenue, Cambridge, MA 02139 | MASSACHUSETTS | AVE 25 Wizard of Oz, Walaford, KS 99912323 | WIZARD OF | 26 Capen Street, Medford, MA | CAPEN | ST 124 Mount Auburn St, Cambridge, Massachusetts 02138 | MOUNT AUBURN | ST 950 Main Street, Worcester, MA 01610 | MAIN | ST
Pprint_Addy — Given a norm_addy
composite type object, returns a pretty print representation of it. Usually used in conjunction with normalize_address.
varchar pprint_addy(
norm_addy in_addy)
;
Given a norm_addy
composite type object, returns a pretty print representation of it. No other data is required aside from what is packaged with the geocoder.
Usually used in conjunction with Normalize_Address.
Pretty print a single address
SELECT pprint_addy(normalize_address('202 East Fremont Street, Las Vegas, Nevada 89101')) As pretty_address; pretty_address --------------------------------------- 202 E Fremont St, Las Vegas, NV 89101
Pretty print address a table of addresses
SELECT address As orig, pprint_addy(normalize_address(address)) As pretty_address FROM addresses_to_geocode; orig | pretty_address -----------------------------------------------------+------------------------------------------- 529 Main Street, Boston MA, 02129 | 529 Main St, Boston MA, 02129 77 Massachusetts Avenue, Cambridge, MA 02139 | 77 Massachusetts Ave, Cambridge, MA 02139 28 Capen Street, Medford, MA | 28 Capen St, Medford, MA 124 Mount Auburn St, Cambridge, Massachusetts 02138 | 124 Mount Auburn St, Cambridge, MA 02138 950 Main Street, Worcester, MA 01610 | 950 Main St, Worcester, MA 01610
Reverse_Geocode — Takes a geometry point in a known spatial ref sys and returns a record containing an array of theoretically possible addresses and an array of cross streets. If include_strnum_range = true, includes the street range in the cross streets.
record Reverse_Geocode(
geometry pt, boolean include_strnum_range=false, geometry[] OUT intpt, norm_addy[] OUT addy, varchar[] OUT street)
;
Takes a geometry point in a known spatial ref and returns a record containing an array of theoretically possible addresses and an array of cross streets. If include_strnum_range = true, includes the street range in the cross streets. include_strnum_range defaults to false if not passed in. Addresses are sorted according to which road a point is closest to so first address is most likely the right one.
Why do we say theoretical instead of actual addresses. The Tiger data doesn't have real addresses, but just street ranges. As such the theoretical address is an interpolated address based on the street ranges. Like for example interpolating one of my addresses returns a 26 Court St. and 26 Court Sq., though there is no such place as 26 Court Sq. This is because a point may be at a corner of 2 streets and thus the logic interpolates along both streets. The logic also assumes addresses are equally spaced along a street, which of course is wrong since you can have a municipal building taking up a good chunk of the street range and the rest of the buildings are clustered at the end.
Note: Hmm this function relies on Tiger data. If you have not loaded data covering the region of this point, then hmm you will get a record filled with NULLS.
Returned elements of the record are as follows:
intpt
is an array of points: These are the center line points on the street closest to the input point. There are as many points as there are addresses.
addy
is an array of norm_addy (normalized addresses): These are an array of possible addresses that fit the input point. The first one in the array is most likely. Generally there should be only one, except in the case when a point is at the corner of 2 or 3 streets, or the point is somewhere on the road and not off to the side.
street
an array of varchar: These are cross streets (or the street) (streets that intersect or are the street the point is projected to be on).
Availability: 2.0.0
Example of a point at the corner of two streets, but closest to one. This is approximate location of MIT: 77 Massachusetts Ave, Cambridge, MA 02139 Note that although we don't have 3 streets, PostgreSQL will just return null for entries above our upper bound so safe to use. This includes street ranges
SELECT pprint_addy(r.addy[1]) As st1, pprint_addy(r.addy[2]) As st2, pprint_addy(r.addy[3]) As st3, array_to_string(r.street, ',') As cross_streets FROM reverse_geocode(ST_GeomFromText('POINT(-71.093902 42.359446)',4269),true) As r; result ------ st1 | st2 | st3 | cross_streets -------------------------------------------+-----+-----+---------------------------------------------- 67 Massachusetts Ave, Cambridge, MA 02139 | | | 67 - 127 Massachusetts Ave,32 - 88 Vassar St
Here we choose not to include the address ranges for the cross streets and picked a location really really close to a corner of 2 streets thus could be known by two different addresses.
SELECT pprint_addy(r.addy[1]) As st1, pprint_addy(r.addy[2]) As st2, pprint_addy(r.addy[3]) As st3, array_to_string(r.street, ',') As cross_str FROM reverse_geocode(ST_GeomFromText('POINT(-71.06941 42.34225)',4269)) As r; result -------- st1 | st2 | st3 | cross_str ---------------------------------+---------------------------------+-----+------------------------ 5 Bradford St, Boston, MA 02118 | 49 Waltham St, Boston, MA 02118 | | Waltham St
For this one we reuse our geocoded example from Geocode and we only want the primary address and at most 2 cross streets.
SELECT actual_addr, lon, lat, pprint_addy((rg).addy[1]) As int_addr1, (rg).street[1] As cross1, (rg).street[2] As cross2 FROM (SELECT address As actual_addr, lon, lat, reverse_geocode( ST_SetSRID(ST_Point(lon,lat),4326) ) As rg FROM addresses_to_geocode WHERE rating > -1) As foo; actual_addr | lon | lat | int_addr1 | cross1 | cross2 -----------------------------------------------------+-----------+----------+-------------------------------------------+-----------------+------------ 529 Main Street, Boston MA, 02129 | -71.07181 | 42.38359 | 527 Main St, Boston, MA 02129 | Medford St | 77 Massachusetts Avenue, Cambridge, MA 02139 | -71.09428 | 42.35988 | 77 Massachusetts Ave, Cambridge, MA 02139 | Vassar St | 26 Capen Street, Medford, MA | -71.12377 | 42.41101 | 9 Edison Ave, Medford, MA 02155 | Capen St | Tesla Ave 124 Mount Auburn St, Cambridge, Massachusetts 02138 | -71.12304 | 42.37328 | 3 University Rd, Cambridge, MA 02138 | Mount Auburn St | 950 Main Street, Worcester, MA 01610 | -71.82368 | 42.24956 | 3 Maywood St, Worcester, MA 01603 | Main St | Maywood Pl
Topology_Load_Tiger — Loads a defined region of tiger data into a PostGIS Topology and transforming the tiger data to spatial reference of the topology and snapping to the precision tolerance of the topology.
text Topology_Load_Tiger(
varchar topo_name, varchar region_type, varchar region_id)
;
Loads a defined region of tiger data into a PostGIS Topology. The faces, nodes and edges are transformed to the spatial reference system of the target topology and points are snapped to the tolerance of the target topology. The created faces, nodes, edges maintain the same ids as the original Tiger data faces, nodes, edges so that datasets can be in the future be more easily reconciled with tiger data. Returns summary details about the process.
This would be useful for example for redistricting data where you require the newly formed polygons to follow the center lines of streets and for the resulting polygons not to overlap.
This function relies on Tiger data as well as the installation of the PostGIS topology module. For more information, refer to Chapter 11, Topologia and Section 2.4.1, “Configurazione”. If you have not loaded data covering the region of interest, then no topology records will be created. This function will also fail if you have not created a topology using the topology functions. |
Most topology validation errors are a result of tolerance issues where after transformation the edges points don't quite line up or overlap. To remedy the situation you may want to increase or lower the precision if you get topology validation failures. |
Required arguments:
topo_name
The name of an existing PostGIS topology to load data into.
region_type
The type of bounding region. Currently only place
and county
are supported. Plan is to have several more. This is the table to look into to define the region bounds. e.g tiger.place
, tiger.county
region_id
This is what TIGER calls the geoid. It is the unique identifier of the region in the table. For place it is the plcidfp
column in tiger.place
. For county it is the cntyidfp
column in tiger.county
Availability: 2.0.0
Create a topology for Boston, Massachusetts in Mass State Plane Feet (2249) with tolerance 0.25 feet and then load in Boston city tiger faces, edges, nodes.
SELECT topology.CreateTopology('topo_boston', 2249, 0.25); createtopology -------------- 15 -- 60,902 ms ~ 1 minute on windows 7 desktop running 9.1 (with 5 states tiger data loaded) SELECT tiger.topology_load_tiger('topo_boston', 'place', '2507000'); -- topology_loader_tiger -- 29722 edges holding in temporary. 11108 faces added. 1875 edges of faces added. 20576 nodes added. 19962 nodes contained in a face. 0 edge start end corrected. 31597 edges added. -- 41 ms -- SELECT topology.TopologySummary('topo_boston'); -- topologysummary-- Topology topo_boston (15), SRID 2249, precision 0.25 20576 nodes, 31597 edges, 11109 faces, 0 topogeoms in 0 layers -- 28,797 ms to validate yeh returned no errors -- SELECT * FROM topology.ValidateTopology('topo_boston'); error | id1 | id2 -------------------+----------+-----------
Create a topology for Suffolk, Massachusetts in Mass State Plane Meters (26986) with tolerance 0.25 meters and then load in Suffolk county tiger faces, edges, nodes.
SELECT topology.CreateTopology('topo_suffolk', 26986, 0.25); -- this took 56,275 ms ~ 1 minute on Windows 7 32-bit with 5 states of tiger loaded -- must have been warmed up after loading boston SELECT tiger.topology_load_tiger('topo_suffolk', 'county', '25025'); -- topology_loader_tiger -- 36003 edges holding in temporary. 13518 faces added. 2172 edges of faces added. 24761 nodes added. 24075 nodes contained in a face. 0 edge start end corrected. 38175 edges added. -- 31 ms -- SELECT topology.TopologySummary('topo_suffolk'); -- topologysummary-- Topology topo_suffolk (14), SRID 26986, precision 0.25 24761 nodes, 38175 edges, 13519 faces, 0 topogeoms in 0 layers -- 33,606 ms to validate -- SELECT * FROM topology.ValidateTopology('topo_suffolk'); error | id1 | id2 -------------------+----------+----------- coincident nodes | 81045651 | 81064553 edge crosses node | 81045651 | 85737793 edge crosses node | 81045651 | 85742215 edge crosses node | 81045651 | 620628939 edge crosses node | 81064553 | 85697815 edge crosses node | 81064553 | 85728168 edge crosses node | 81064553 | 85733413
Set_Geocode_Setting — Sets a setting that affects behavior of geocoder functions.
text Set_Geocode_Setting(
text setting_name, text setting_value)
;
Sets value of specific setting stored in tiger.geocode_settings
table. Settings allow you to toggle debugging of functions. Later plans will be to control rating with settings. Current list of settings are listed in Get_Geocode_Setting.
Availability: 2.1.0
If you run Geocode when this function is true, the NOTICE log will output timing and queries.
SELECT set_geocode_setting('debug_geocode_address', 'true') As result; result --------- true
Table of Contents
The functions given below are spatial aggregate functions provided with PostGIS that can be used just like any other sql aggregate function such as sum, average.
The functions given below are PostGIS functions that conform to the SQL/MM 3 standard
SQL-MM defines the default SRID of all geometry constructors as 0. PostGIS uses a default SRID of -1. |
The functions and operators given below are PostGIS functions/operators that take as input or return as output a geography data type object.
Functions with a (T) are not native geodetic functions, and use a ST_Transform call to and from geometry to do the operation. As a result, they may not behave as expected when going over dateline, poles, and for large geometries or geometry pairs that cover more than one UTM zone. Basic transform - (favoring UTM, Lambert Azimuthal (North/South), and falling back on mercator in worst case scenario) |
The functions and operators given below are PostGIS functions/operators that take as input or return as output a raster data type object. Listed in alphabetical order.
The functions given below are PostGIS functions that take as input or return as output a set of or single geometry_dump or geomval data type object.
The functions given below are PostGIS functions that take as input or return as output the box* family of PostGIS spatial types. The box family of types consists of box2d, and box3d
The functions given below are PostGIS functions that do not throw away the Z-Index.
The functions given below are PostGIS functions that can use CIRCULARSTRING, CURVEPOLYGON, and other curved geometry types
The functions given below are PostGIS functions that can use POLYHEDRALSURFACE, POLYHEDRALSURFACEM geometries
Below is an alphabetical listing of spatial specific functions in PostGIS and the kinds of spatial types they work with or OGC/SQL compliance they try to conform to.
The functions given below are PostGIS functions that were added or enhanced.
Functions new in PostGIS 2.3
The functions given below are PostGIS functions that are enhanced in PostGIS 2.3.
The functions given below are PostGIS functions that were added or enhanced.
postgis_sfcgal now can be installed as an extension using CREATE EXTENSION postgis_sfcgal; |
PostGIS 2.2.0: Tiger Geocoder upgraded to work with TIGER 2015 data. |
address_standardizer, address_standardizer_data_us extensions for standardizing address data refer to Chapter 12, Address Standardizer for details. |
Many functions in topology rewritten as C functions for increased performance. |
Functions new in PostGIS 2.2
The functions given below are PostGIS functions that are enhanced in PostGIS 2.2.
The functions given below are PostGIS functions that have possibly breaking changes in PostGIS 2.2. If you use any of these, you may need to check your existing code.
The functions given below are PostGIS functions that were added or enhanced.
More Topology performance Improvements. Please refer to Chapter 11, Topologia for more details. |
Bug fixes (particularly with handling of out-of-band rasters), many new functions (often shortening code you have to write to accomplish a common task) and massive speed improvements to raster functionality. Refer to Chapter 9, Raster Reference for more details. |
PostGIS 2.1.0: Tiger Geocoder upgraded to work with TIGER 2012 census data. |
Functions new in PostGIS 2.1
The functions given below are PostGIS functions that are enhanced in PostGIS 2.1.
The functions given below are PostGIS functions that have possibly breaking changes in PostGIS 2.1. If you use any of these, you may need to check your existing code.
The functions given below are PostGIS functions that were added, enhanced, or have Section 14.11.7, “PostGIS Functions changed behavior in 2.0” breaking changes in 2.0 releases.
New geometry types: TIN and Polyhedral surfaces was introduced in 2.0
Greatly improved support for Topology. Please refer to Chapter 11, Topologia for more details. |
In PostGIS 2.0, raster type and raster functionality has been integrated. There are way too many new raster functions to list here and all are new so
please refer to Chapter 9, Raster Reference for more details of the raster functions available. Earlier pre-2.0 versions had raster_columns/raster_overviews as real tables. These were changed to views before release. Functions such as |
Tiger Geocoder upgraded to work with TIGER 2010 census data and now included in the core PostGIS documentation. A reverse geocoder function was also added. Please refer to Section 13.1, “Tiger Geocoder” for more details. |
The functions given below are PostGIS functions that are enhanced in PostGIS 2.0.
The functions given below are PostGIS functions that have changed behavior in PostGIS 2.0 and may require application changes.
Most deprecated functions have been removed. These are functions that haven't been documented since 1.2 or some internal functions that were never documented. If you are using a function that you don't see documented, it's probably deprecated, about to be deprecated, or internal and should be avoided. If you have applications or tools that rely on deprecated functions, please refer to Q: 3.2 for more details. |
Bounding boxes of geometries have been changed from float4 to double precision (float8). This has an impact on answers you get using bounding box operators and casting of bounding boxes to geometries. E.g ST_SetSRID(abbox) will often return a different more accurate answer in PostGIS 2.0+ than it did in prior versions which may very well slightly change answers to view port queries. |
The arguments hasnodata was replaced with exclude_nodata_value which has the same meaning as the older hasnodata but clearer in purpose. |
The functions given below are PostGIS functions that were introduced or enhanced in this minor release.
The functions given below are PostGIS functions that were introduced or enhanced in the 1.4 release.
The functions given below are PostGIS functions that were introduced in the 1.3 release.
Segnalare efficacemente i bug è una modalità fondamentale per aiutare lo sviluppo di PostGIS. Il bug report più efficiente è quello che permette agli sviluppatori PostGIS di replicarlo, quindi idealmente contiene uno script che lo rende evidente e un elenco di informazioni sull'ambiente in cui si verifica. Le informazioni indispensabili possono essere estratte eseguendo SELECT postgis_full_version()
[per postgis] and SELECT version()
[per postgresql].
Se non si sta usando l'ultima release, vale la pena leggerne il release changelog, per scoprire se il bug in questione è già stato risolto.
Usare il PostGIS bug tracker assicura che le segnalazioni non vadano perse, e permette di restare aggiornati sulla sua gestione. Prima di riportare un nuovo bug, per favore esaminate il database dei bug per vedere se è già stato segnalato, nel qual caso è opportuno aggiungere ad esso nuove informazioni.
Consigliamo la lettura del paper di Simon Tatham su How to Report Bugs Effectively (Come riportare bug in modo efficiente) prima di compilare un nuovo bug report.
La documentazione deve riflettere accuratamente le funzionalità e il comportamento del software. Se così non fosse, potrebbe essere dovuto a un bug del software o a un errore o mancanza della documentazione.
I problemi della documentazione possono essere riportati nel bug tracker di PostGIS.
Se la revisione è minore, basta descriverla in una nuova issue del bug tracker, specificando la sua posizione nella documentazione.
Se i cambiamenti sono più estesi, è preferibile mandare una patch per Subversion. In Unix, si tratta di un processo in 4 fasi (assumendo di avere Subversion già installato):
Eseguire il checkout del trunk di PostGIS da Subversion. In Unix, digitare:
svn checkout http://svn.osgeo.org/postgis/trunk/
I file saranno salvati nella cartella ./trunk
Apportare le modifiche alla documentazione con l'editor di testo preferito. In Unix, ad esempio, digitare:
vim trunk/doc/postgis.xml
Nota: la documentazione è scritta in DocBook XML, non in HTML, quindi se non si ha familiarità con esso conviene seguire lo stile della documentazione esistente.
Produrre una patch, cioè un file contenente le differenze rispetto alla copia master della documentazione. In Unix, digitare:
svn diff trunk/doc/postgis.xml > doc.patch
Allegare la patch a una nuova issue nel bug tracker.
Table of Contents
Release date: 2016/01/06
This is a bug fix and performance improvement release.
#2232, avoid accumulated error in SVG rounding
#3321, Fix performance regression in topology loading
#3329, Fix robustness regression in TopoGeo_addPoint
#3349, Fix installation path of postgis_topology scripts
#3351, set endnodes isolation on ST_RemoveIsoEdge (and lwt_RemIsoEdge)
#3355, geography ST_Segmentize has geometry bbox
#3359, Fix toTopoGeom loss of low-id primitives from TopoGeometry definition
#3360, _raster_constraint_info_scale invalid input syntax
#3375, crash in repeated point removal for collection(point)
#3378, Fix handling of hierarchical TopoGeometries in presence of multiple topologies
#3380, #3402, Decimate lines on topology load
#3388, #3410, Fix missing end-points in ST_Removepoints
#3389, Buffer overflow in lwgeom_to_geojson
#3390, Compilation under Alpine Linux 3.2 gives an error when compiling the postgis and postgis_topology extension
#3393, ST_Area NaN for some polygons
#3401, Improve ST_Split robustness on 32bit systems
#3404, ST_ClusterWithin crashes backend
#3407, Fix crash on splitting a face or an edge defining multiple TopoGeometry objects
#3411, Clustering functions not using spatial index
#3412, Improve robustness of snapping step in TopoGeo_addLinestring
#3415, Fix OSX 10.9 build under pkgsrc
Fix memory leak in lwt_ChangeEdgeGeom [liblwgeom]
Release date: 2015/10/07
This is a new feature release, with new functions, improved performance, and other goodies.
Topology API in liblwgeom (Sandro Santilli / Regione Toscana - SITA)
New lwgeom_unaryunion method in liblwgeom
New lwgeom_linemerge method in liblwgeom
New lwgeom_is_simple method in liblwgeom
#3169, Add SFCGAL 1.1 support: add ST_3DDifference, ST_3DUnion, ST_Volume, ST_MakeSolid, ST_IsSolid (Vincent Mora / Oslandia)
#3169, ST_ApproximateMedialAxis (Sandro Santilli)
ST_CPAWithin (Sandro Santilli / Boundless)
Add |=| operator with CPA semantic and KNN support with PgSQL 9.5+ (Sandro Santilli / Boundless)
#3131, KNN support for the geography type (Paul Ramsey / CartoDB)
#3023, ST_ClusterIntersecting / ST_ClusterWithin (Dan Baston)
#2703, Exact KNN results for all geometry types, aka "KNN re-check" (Paul Ramsey / CartoDB)
#1137, Allow a tolerance value in ST_RemoveRepeatedPoints (Paul Ramsey / CartoDB)
#3062, Allow passing M factor to ST_Scale (Sandro Santilli / Boundless)
#3139, ST_BoundingDiagonal (Sandro Santilli / Boundless)
#3129, ST_IsValidTrajectory (Sandro Santilli / Boundless)
#3128, ST_ClosestPointOfApproach (Sandro Santilli / Boundless)
#3152, ST_DistanceCPA (Sandro Santilli / Boundless)
Canonical output for index key types
ST_SwapOrdinates (Sandro Santilli / Boundless)
#2918, Use GeographicLib functions for geodetics (Mike Toews)
#3074, ST_Subdivide to break up large geometry (Paul Ramsey / CartoDB)
#3040, KNN GiST index based centroid (<<->>) n-D distance operators (Sandro Santilli / Boundless)
Interruptibility API for liblwgeom (Sandro Santilli / CartoDB)
#2939, ST_ClipByBox2D (Sandro Santilli / CartoDB)
#2247, ST_Retile and ST_CreateOverview: in-db raster overviews creation (Sandro Santilli / Vizzuality)
#899, -m shp2pgsql attribute names mapping -m switch (Regina Obe / Sandro Santilli)
#1678, Added GUC postgis.gdal_datapath to specify GDAL config variable GDAL_DATA
#2843, Support reprojection on raster import (Sandro Santilli / Vizzuality)
#2349, Support for encoded_polyline input/output (Kashif Rasul)
#2159, report libjson version from postgis_full_version()
#2770, ST_MemSize(raster)
Add postgis_noop(raster)
Added missing variants of ST_TPI(), ST_TRI() and ST_Roughness()
Added GUC postgis.gdal_enabled_drivers to specify GDAL config variable GDAL_SKIP
Added GUC postgis.enable_outdb_rasters to enable access to rasters with out-db bands
#2387, address_standardizer extension as part of PostGIS (Stephen Woodbridge / imaptools.com, Walter Sinclair, Regina Obe)
#2816, address_standardizer_data_us extension provides reference lex,gaz,rules for address_standardizer (Stephen Woodbridge / imaptools.com, Walter Sinclair, Regina Obe)
#2341, New mask parameter for ST_MapAlgebra
#2397, read encoding info automatically in shapefile loader
#2430, ST_ForceCurve
#2565, ST_SummaryStatsAgg()
#2567, ST_CountAgg()
#2632, ST_AsGML() support for curved features
#2652, Add --upgrade-path switch to run_test.pl
#2754, sfcgal wrapped as an extension
#2227, Simplification with Visvalingam-Whyatt algorithm ST_SimplifyVW, ST_SetEffectiveArea (Nicklas Avén)
Functions to encode and decode TWKB ST_AsTWKB, ST_GeomFromTWKB (Paul Ramsey / Nicklas Avén / CartoDB)
#3223, Add memcmp short-circuit to ST_Equals (Daniel Baston)
#3227, Tiger geocoder upgraded to support Tiger 2015 census
#2278, Make liblwgeom compatible between minor releases
#897, ST_AsX3D support for GeoCoordinates and systems "GD" "WE" ability to flip x/y axis (use option = 2, 3)
ST_Split: allow splitting lines by multilines, multipoints and (multi)polygon boundaries
#3070, Simplify geometry type constraint
#2839, Implement selectivity estimator for functional indexes, speeding up spatial queries on raster tables. (Sandro Santilli / Vizzuality)
#2361, Added spatial_index column to raster_columns view
#2390, Testsuite for pgsql2shp
#2527, Added -k flag to raster2pgsql to skip checking that band is NODATA
#2616, Reduce text casts during topology building and export
#2717, support startpoint, endpoint, pointn, numpoints for compoundcurve
#2747, Add support for GDAL 2.0
#2754, SFCGAL can now be installed with CREATE EXTENSION (Vincent Mora @ Oslandia)
#2828, Convert ST_Envelope(raster) from SQL to C
#2829, Shortcut ST_Clip(raster) if geometry fully contains the raster and no NODATA specified
#2906, Update tiger geocoder to handle tiger 2014 data
#3048, Speed up geometry simplification (J.Santana @ CartoDB)
#3092, Slow performance of geometry_columns with many tables
Release date: 2015-07-07
This is a critical bug fix release.
#3159, do not force a bbox cache on ST_Affine
#3018, GROUP BY geography sometimes returns duplicate rows
#3048, shp2pgsql - illegal number format when specific system locale set
#3094, Malformed GeoJSON inputs crash backend
#3104, st_asgml introduces random characters in ID field
#3155, Remove liblwgeom.h on make uninstall
#3177, gserialized_is_empty cannot handle nested empty cases
Fix crash in ST_LineLocatePoint
Release date: 2015-03-30
This is a critical bug fix release.
Release date: 2015-03-20
This is a bug fix and performance improvement release.
#3000, Ensure edge splitting and healing algorithms use indexes
#3048, Speed up geometry simplification (J.Santana @ CartoDB)
#3050, Speep up geometry type reading (J.Santana @ CartoDB)
#2941, allow geography columns with SRID other than 4326
#3069, small objects getting inappropriately fluffed up w/ boxes
#3068, Have postgis_typmod_dims return NULL for unconstrained dims
#3061, Allow duplicate points in JSON, GML, GML ST_GeomFrom* functions
#3058, Fix ND-GiST picksplit method to split on the best plane
#3052, Make operators <-> and <#> available for PostgreSQL < 9.1
#3045, Fix dimensionality confusion in &&& operator
#3016, Allow unregistering layers of corrupted topologies
#3015, Avoid exceptions from TopologySummary
#3020, ST_AddBand out-db bug where height using width value
#3031, Allow restore of Geometry(Point) tables dumped with empties in them
Release date: 2014-12-18
This is a bug fix and performance improvement release.
#2933, Speedup construction of large multi-geometry objects
#2947, Fix memory leak in lwgeom_make_valid for single-component collection input
#2949, Fix memory leak in lwgeom_mindistance2d for curve input
#2931, BOX representation is case sensitive
#2942, PostgreSQL 9.5 support
#2953, 2D stats not generated when Z/M values are extreme
#3009, Geography cast may effect underlying tuple
Release date: 2014-09-10
This is a bug fix and performance improvement release.
#2745, Speedup ST_Simplify calls against points
#2747, Support for GDAL 2.0
#2749, Make rtpostgis_upgrade_20_21.sql ACID
#2811, Do not specify index names when loading shapefiles/rasters
#2829, Shortcut ST_Clip(raster) if geometry fully contains the raster and no NODATA specified
#2895, Raise cost of ST_ConvexHull(raster) to 300 for better query plans
#2605, armel: _ST_Covers() returns true for point in hole
#2911, Fix output scale on ST_Rescale/ST_Resample/ST_Resize of rasters with scale 1/-1 and offset 0/0.
Fix crash in ST_Union(raster)
#2704, ST_GeomFromGML() does not work properly with array of gml:pos (Even Roualt)
#2708, updategeometrysrid doesn't update srid check when schema not specified. Patch from Marc Jansen
#2720, lwpoly_add_ring should update maxrings after realloc
#2759, Fix postgis_restore.pl handling of multiline object comments embedding sql comments
#2774, fix undefined behavior in ptarray_calculate_gbox_geodetic
Fix potential memory fault in ST_MakeValid
#2784, Fix handling of bogus argument to --with-sfcgal
#2772, Premature memory free in RASTER_getBandPath (ST_BandPath)
#2755, Fix regressions tests against all versions of SFCGAL
#2775, lwline_from_lwmpoint leaks memory
#2802, ST_MapAlgebra checks for valid callback function return value
#2803, ST_MapAlgebra handles no userarg and STRICT callback function
#2834, ST_Estimated_Extent and mixedCase table names (regression bug)
#2845, Bad geometry created from ST_AddPoint
#2870, Binary insert into geography column results geometry being inserted
#2872, make install builds documentation (Greg Troxell)
#2819, find isfinite or replacement on Centos5 / Solaris
#2899, geocode limit 1 not returning best answer (tiger geocoder)
#2903, Unable to compile on FreeBSD
#2927 reverse_geocode not filling in direction prefix (tiger geocoder) get rid of deprecated ST_Line_Locate_Point called
Release date: 2014/05/13
This is a bug fix and security release.
Starting with this version offline raster access and use of GDAL drivers are disabled by default.
An environment variable is introduced to allow for enabling specific GDAL drivers: POSTGIS_GDAL_ENABLED_DRIVERS. By default, all GDAL drivers are disabled
An environment variable is introduced to allow for enabling out-db raster bands: POSTGIS_ENABLE_OUTDB_RASTERS. By default, out-db raster bands are disabled
The environment variables must be set for the PostgreSQL process, and determines the behavior of the whole cluster.
Release date: 2014/03/31
This is a bug fix release, addressing issues that have been filed since the 2.1.1 release.
#2666, Error out at configure time if no SQL preprocessor can be found
#2534, st_distance returning incorrect results for large geographies
#2539, Check for json-c/json.h presence/usability before json/json.h
#2543, invalid join selectivity error from simple query
#2546, GeoJSON with string coordinates parses incorrectly
#2547, Fix ST_Simplify(TopoGeometry) for hierarchical topogeoms
#2552, Fix NULL raster handling in ST_AsPNG, ST_AsTIFF and ST_AsJPEG
#2555, Fix parsing issue of range arguments of ST_Reclass
#2556, geography ST_Intersects results depending on insert order
#2580, Do not allow installing postgis twice in the same database
#2589, Remove use of unnecessary void pointers
#2607, Cannot open more than 1024 out-db files in one process
#2610, Ensure face splitting algorithm uses the edge index
#2615, EstimatedExtent (and hence, underlying stats) gathering wrong bbox
#2619, Empty rings array in GeoJSON polygon causes crash
#2634, regression in sphere distance code
#2638, Geography distance on M geometries sometimes wrong
#2648, #2653, Fix topology functions when "topology" is not in search_path
#2654, Drop deprecated calls from topology
#2655, Let users without topology privileges call postgis_full_version()
#2674, Fix missing operator = and hash_raster_ops opclass on raster
#2675, #2534, #2636, #2634, #2638, Geography distance issues with tree optimization
Release date: 2013/11/06
This is a bug fix release, addressing issues that have been filed since the 2.1.0 release.
#2514, Change raster license from GPL v3+ to v2+, allowing distribution of PostGIS Extension as GPLv2.
#2396, Make regression tests more endian-agnostic
#2434, Fix ST_Intersection(geog,geog) regression in rare cases
#2454, Fix behavior of ST_PixelAsXXX functions regarding exclude_nodata_value parameter
#2489, Fix upgrades from 2.0 leaving stale function signatures
#2525, Fix handling of SRID in nested collections
#2449, Fix potential infinite loop in index building
#2493, Fix behavior of ST_DumpValues when passed an empty raster
#2502, Fix postgis_topology_scripts_installed() install schema
#2504, Fix segfault on bogus pgsql2shp call
#2512, Support for foreign tables and materialized views in raster_columns and raster_overviews
Release date: 2013/08/17
This is a minor release addressing both bug fixes and performance and functionality enhancements addressing issues since 2.0.3 release. If you are upgrading from 2.0+, only a soft upgrade is required. If you are upgrading from 1.5 or earlier, a hard upgrade is required.
#1653, Removed srid parameter from ST_Resample(raster) and variants with reference raster no longer apply reference raster's SRID.
#1962 ST_Segmentize - As a result of the introduction of geography support, The construct: SELECT ST_Segmentize('LINESTRING(1 2, 3 4)',0.5);
will result in ambiguous function error
#2026, ST_Union(raster) now unions all bands of all rasters
#2089, liblwgeom: lwgeom_set_handlers replaces lwgeom_init_allocators.
#2150, regular_blocking is no longer a constraint. column of same name in raster_columns now checks for existance of spatially_unique and coverage_tile constraints
ST_Intersects(raster, geometry) behaves in the same manner as ST_Intersects(geometry, raster).
point variant of ST_SetValue(raster) previously did not check SRID of input geometry and raster.
ST_Hillshade parameters azimuth and altitude are now in degrees instead of radians.
ST_Slope and ST_Aspect return pixel values in degrees instead of radians.
#2104, ST_World2RasterCoord, ST_World2RasterCoordX and ST_World2RasterCoordY renamed to ST_WorldToRasterCoord, ST_WorldToRasterCoordX and ST_WorldToRasterCoordY. ST_Raster2WorldCoord, ST_Raster2WorldCoordX and ST_Raster2WorldCoordY renamed to ST_RasterToWorldCoord, ST_RasterToWorldCoordX and ST_RasterToWorldCoordY
ST_Estimated_Extent renamed to ST_EstimatedExtent
ST_Line_Interpolate_Point renamed to ST_LineInterpolatePoint
ST_Line_Substring renamed to ST_LineSubstring
ST_Line_Locate_Point renamed to ST_LineLocatePoint
ST_Force_XXX renamed to ST_ForceXXX
ST_MapAlgebraFctNgb and 1 and 2 raster variants of ST_MapAlgebraFct. Use ST_MapAlgebra instead
1 and 2 raster variants of ST_MapAlgebraExpr. Use expression variants of ST_MapAlgebra instead
- Refer to http://postgis.net/docs/manual-2.1/PostGIS_Special_Functions_Index.html#NewFunctions_2_1 for complete list of new functions
#310, ST_DumpPoints converted to a C function (Nathan Wagner) and much faster
#739, UpdateRasterSRID()
#945, improved join selectivity, N-D selectivity calculations, user accessible selectivity and stats reader functions for testing (Paul Ramsey / OpenGeo)
toTopoGeom with TopoGeometry sink (Sandro Santilli / Vizzuality)
clearTopoGeom (Sandro Santilli / Vizzuality)
ST_Segmentize(geography) (Paul Ramsey / OpenGeo)
ST_DelaunayTriangles (Sandro Santilli / Vizzuality)
ST_NearestValue, ST_Neighborhood (Bborie Park / UC Davis)
ST_PixelAsPoint, ST_PixelAsPoints (Bborie Park / UC Davis)
ST_PixelAsCentroid, ST_PixelAsCentroids (Bborie Park / UC Davis)
ST_Raster2WorldCoord, ST_World2RasterCoord (Bborie Park / UC Davis)
Additional raster/raster spatial relationship functions (ST_Contains, ST_ContainsProperly, ST_Covers, ST_CoveredBy, ST_Disjoint, ST_Overlaps, ST_Touches, ST_Within, ST_DWithin, ST_DFullyWithin) (Bborie Park / UC Davis)
Added array variants of ST_SetValues() to set many pixel values of a band in one call (Bborie Park / UC Davis)
#1293, ST_Resize(raster) to resize rasters based upon width/height
#1627, package tiger_geocoder as a PostgreSQL extension
#1643, #2076, Upgrade tiger geocoder to support loading tiger 2011 and 2012 (Regina Obe / Paragon Corporation) Funded by Hunter Systems Group
GEOMETRYCOLLECTION support for ST_MakeValid (Sandro Santilli / Vizzuality)
#1709, ST_NotSameAlignmentReason(raster, raster)
#1818, ST_GeomFromGeoHash and friends (Jason Smith (darkpanda))
#1856, reverse geocoder rating setting for prefer numbered highway name
ST_PixelOfValue (Bborie Park / UC Davis)
Casts to/from PostgreSQL geotypes (point/path/polygon).
Added geomval array variant of ST_SetValues() to set many pixel values of a band using a set of geometries and corresponding values in one call (Bborie Park / UC Davis)
ST_Tile(raster) to break up a raster into tiles (Bborie Park / UC Davis)
#1895, new r-tree node splitting algorithm (Alex Korotkov)
#2011, ST_DumpValues to output raster as array (Bborie Park / UC Davis)
#2018, ST_Distance support for CircularString, CurvePolygon, MultiCurve, MultiSurface, CompoundCurve
#2030, n-raster (and n-band) ST_MapAlgebra (Bborie Park / UC Davis)
#2193, Utilize PAGC parser as drop in replacement for tiger normalizer (Steve Woodbridge, Regina Obe)
#2210, ST_MinConvexHull(raster)
lwgeom_from_geojson in liblwgeom (Sandro Santilli / Vizzuality)
#1687, ST_Simplify for TopoGeometry (Sandro Santilli / Vizzuality)
#2228, TopoJSON output for TopoGeometry (Sandro Santilli / Vizzuality)
#2123, ST_FromGDALRaster
#613, ST_SetGeoReference with numerical parameters instead of text
#2276, ST_AddBand(raster) variant for out-db bands
#2280, ST_Summary(raster)
#2163, ST_TPI for raster (Nathaniel Clay)
#2164, ST_TRI for raster (Nathaniel Clay)
#2302, ST_Roughness for raster (Nathaniel Clay)
#2290, ST_ColorMap(raster) to generate RGBA bands
#2254, Add SFCGAL backend support. (Backend selection throught postgis.backend var) Functions available both throught GEOS or SFCGAL: ST_Intersects, ST_3DIntersects, ST_Intersection, ST_Area, ST_Distance, ST_3DDistance New functions available only with SFCGAL backend: ST_3DIntersection, ST_Tesselate, ST_3DArea, ST_Extrude, ST_ForceLHR ST_Orientation, ST_Minkowski, ST_StraightSkeleton postgis_sfcgal_version New function available in PostGIS: ST_ForceSFS (Olivier Courtin and Hugo Mercier / Oslandia)
For detail of new functions and function improvements, please refer to Section 14.11.4, “PostGIS Functions new or enhanced in 2.1”.
Much faster raster ST_Union, ST_Clip and many more function additions operations
For geometry/geography better planner selectivity and a lot more functions.
#823, tiger geocoder: Make loader_generate_script download portion less greedy
#826, raster2pgsql no longer defaults to padding tiles. Flag -P can be used to pad tiles
#1363, ST_AddBand(raster, ...) array version rewritten in C
#1364, ST_Union(raster, ...) aggregate function rewritten in C
#1655, Additional default values for parameters of ST_Slope
#1661, Add aggregate variant of ST_SameAlignment
#1719, Add support for Point and GeometryCollection ST_MakeValid inputs
#1780, support ST_GeoHash for geography
#1796, Big performance boost for distance calculations in geography
#1802, improved function interruptibility.
#1823, add parameter in ST_AsGML to use id column for GML 3 output (become mandatory since GML 3.2.1)
#1856, tiger geocoder: reverse geocoder rating setting for prefer numbered highway name
#1938, Refactor basic ST_AddBand to add multiple new bands in one call
#1978, wrong answer when calculating length of a closed circular arc (circle)
#1989, Preprocess input geometry to just intersection with raster to be clipped
#2021, Added multi-band support to ST_Union(raster, ...) aggregate function
#2006, better support of ST_Area(geography) over poles and dateline
#2065, ST_Clip(raster, ...) now a C function
#2069, Added parameters to ST_Tile(raster) to control padding of tiles
#2078, New variants of ST_Slope, ST_Aspect and ST_HillShade to provide solution to handling tiles in a coverage
#2097, Added RANGE uniontype option for ST_Union(raster)
#2105, Added ST_Transform(raster) variant for aligning output to reference raster
#2119, Rasters passed to ST_Resample(), ST_Rescale(), ST_Reskew(), and ST_SnapToGrid() no longer require an SRID
#2141, More verbose output when constraints fail to be added to a raster column
#2143, Changed blocksize constraint of raster to allow multiple values
#2148, Addition of coverage_tile constraint for raster
#2149, Addition of spatially_unique constraint for raster
TopologySummary output now includes unregistered layers and a count of missing TopoGeometry objects from their natural layer.
ST_HillShade(), ST_Aspect() and ST_Slope() have one new optional parameter to interpolate NODATA pixels before running the operation.
Point variant of ST_SetValue(raster) is now a wrapper around geomval variant of ST_SetValues(rast).
Proper support for raster band's isnodata flag in core API and loader.
Additional default values for parameters of ST_Aspect and ST_HillShade
#2178, ST_Summary now advertises presence of known srid with an [S] flag
#2202, Make libjson-c optional (--without-json configure switch)
#2213, Add support libjson-c 0.10+
#2231, raster2pgsql supports user naming of filename column with -n
#2200, ST_Union(raster, uniontype) unions all bands of all rasters
#2264, postgis_restore.pl support for restoring into databases with postgis in a custom schema
#2244, emit warning when changing raster's georeference if raster has out-db bands
#2222, add parameter OutAsIn to flag whether ST_AsBinary should return out-db bands as in-db bands
#1839, handling of subdatasets in GeoTIFF in raster2pgsql.
#1840, fix logic of when to compute # of tiles in raster2pgsql.
#1870, align the docs and actual behavior of raster's ST_Intersects
#1872, fix ST_ApproxSummarystats to prevent division by zero
#1875, ST_SummaryStats returns NULL for all parameters except count when count is zero
#1932, fix raster2pgsql of syntax for index tablespaces
#1936, ST_GeomFromGML on CurvePolygon causes server crash
#1939, remove custom data types: summarystats, histogram, quantile, valuecount
#1951, remove crash on zero-length linestrings
#1957, ST_Distance to a one-point LineString returns NULL
#1976, Geography point-in-ring code overhauled for more reliability
#1981, cleanup of unused variables causing warnings with gcc 4.6+
#1996, support POINT EMPTY in GeoJSON output
#2062, improve performance of distance calculations
#2057, Fixed linking issue for raster2psql to libpq
#2077, Fixed incorrect values returning from ST_Hillshade()
#2019, ST_FlipCoordinates does not update bbox
#2100, ST_AsRaster may not return raster with specified pixel type
#2126, Better handling of empty rasters from ST_ConvexHull()
#2165, ST_NumPoints regression failure with CircularString
#2168, ST_Distance is not always commutative
#2182, Fix issue with outdb rasters with no SRID and ST_Resize
#2188, Fix function parameter value overflow that caused problems when copying data from a GDAL dataset
#2198, Fix incorrect dimensions used when generating bands of out-db rasters in ST_Tile()
#2201, ST_GeoHash wrong on boundaries
#2203, Changed how rasters with unknown SRID and default geotransform are handled when passing to GDAL Warp API
#2215, Fixed raster exclusion constraint for conflicting name of implicit index
#2251, Fix bad dimensions when rescaling rasters with default geotransform matrix
#2133, Fix performance regression in expression variant of ST_MapAlgebra
#2257, GBOX variables not initialized when testing with empty geometries
#2271, Prevent parallel make of raster
#2282, Fix call to undefined function nd_stats_to_grid() in debug mode
#2307, ST_MakeValid outputs invalid geometries
#2309, Remove confusing INFO message when trying to get SRS info
#2336, FIPS 20 (KS) causes wildcard expansion to wget all files
#2348, Provide raster upgrade path for 2.0 to 2.1
#2351, st_distance between geographies wrong
#2359, Fix handling of schema name when adding overview constraints
#2371, Support GEOS versions with more than 1 digit in micro
#2383, Remove unsafe use of \' from raster warning message
#2384, Incorrect variable datatypes for ST_Neighborhood
#2111, Raster bands can only reference the first 256 bands of out-db rasters
Release date: 2014/03/31
This is a bug fix release, addressing issues that have been filed since the 2.0.4 release. If you are using PostGIS 2.0+ a soft upgrade is required. For users of PostGIS 1.5 or below, a hard upgrade is required.
#2494, avoid memcpy in GIST index
#2502, Fix postgis_topology_scripts_installed() install schema
#2504, Fix segfault on bogus pgsql2shp call
#2528, Fix memory leak in ST_Split / lwline_split_by_line
#2532, Add missing raster/geometry commutator operators
#2533, Remove duplicated signatures
#2552, Fix NULL raster handling in ST_AsPNG, ST_AsTIFF and ST_AsJPEG
#2555, Fix parsing issue of range arguments of ST_Reclass
#2589, Remove use of unnecessary void pointers
#2607, Cannot open more than 1024 out-db files in process
#2610, Ensure face splitting algorithm uses the edge index
#2619, Empty ring array in GeoJSON polygon causes crash
#2638, Geography distance on M geometries sometimes wrong
##2514, Change raster license from GPL v3+ to v2+, allowing distribution of PostGIS Extension as GPLv2.
Release date: 2013/09/06
This is a bug fix release, addressing issues that have been filed since the 2.0.3 release. If you are using PostGIS 2.0+ a soft upgrade is required. For users of PostGIS 1.5 or below, a hard upgrade is required.
#2110, Equality operator between EMPTY and point on origin
Allow adding points at precision distance with TopoGeo_addPoint
#1968, Fix missing edge from toTopoGeom return
#2165, ST_NumPoints regression failure with CircularString
#2168, ST_Distance is not always commutative
#2186, gui progress bar updates too frequent
#2201, ST_GeoHash wrong on boundaries
#2257, GBOX variables not initialized when testing with empty geometries
#2271, Prevent parallel make of raster
#2267, Server crash from analyze table
#2277, potential segfault removed
#2307, ST_MakeValid outputs invalid geometries
#2351, st_distance between geographies wrong
#2359, Incorrect handling of schema for overview constraints
#2371, Support GEOS versions with more than 1 digit in micro
#2372, Cannot parse space-padded KML coordinates
Fix build with systemwide liblwgeom installed
#2383, Fix unsafe use of \' in warning message
#2410, Fix segmentize of collinear curve
#2412, ST_LineToCurve support for lines with less than 4 vertices
#2415, ST_Multi support for COMPOUNDCURVE and CURVEPOLYGON
#2420, ST_LineToCurve: require at least 8 edges to define a full circle
#2423, ST_LineToCurve: require all arc edges to form the same angle
#2424, ST_CurveToLine: add support for COMPOUNDCURVE in MULTICURVE
#2427, Make sure to retain first point of curves on ST_CurveToLine
#2269, Avoid uselessly detoasting full geometries on ANALYZE
#2111, Raster bands can only reference the first 256 bands of out-db rasters
Release date: 2013/03/01
This is a bug fix release, addressing issues that have been filed since the 2.0.2 release. If you are using PostGIS 2.0+ a soft upgrade is required. For users of PostGIS 1.5 or below, a hard upgrade is required.
#2126, Better handling of empty rasters from ST_ConvexHull()
#2134, Make sure to process SRS before passing it off to GDAL functions
Fix various memory leaks in liblwgeom
#2173, Fix robustness issue in splitting a line with own vertex also affecting topology building (#2172)
#2174, Fix usage of wrong function lwpoly_free()
#2176, Fix robustness issue with ST_ChangeEdgeGeom
#2184, Properly copy topologies with Z value
postgis_restore.pl support for mixed case geometry column name in dumps
#2188, Fix function parameter value overflow that caused problems when copying data from a GDAL dataset
#2216, More memory errors in MultiPolygon GeoJSON parsing (with holes)
Fix Memory leak in GeoJSON parser
#2141, More verbose output when constraints fail to be added to a raster column
Speedup ST_ChangeEdgeGeom
Release date: 2012/12/03
This is a bug fix release, addressing issues that have been filed since the 2.0.1 release.
#1287, Drop of "gist_geometry_ops" broke a few clients package of legacy_gist.sql for these cases
#1391, Errors during upgrade from 1.5
#1828, Poor selectivity estimate on ST_DWithin
#1838, error importing tiger/line data
#1869, ST_AsBinary is not unique added to legacy_minor/legacy.sql scripts
#1885, Missing field from tabblock table in tiger2010 census_loader.sql
#1891, Use LDFLAGS environment when building liblwgeom
#1900, Fix pgsql2shp for big-endian systems
#1932, Fix raster2pgsql for invalid syntax for setting index tablespace
#1936, ST_GeomFromGML on CurvePolygon causes server crash
#1955, ST_ModEdgeHeal and ST_NewEdgeHeal for doubly connected edges
#1957, ST_Distance to a one-point LineString returns NULL
#1976, Geography point-in-ring code overhauled for more reliability
#1978, wrong answer calculating length of closed circular arc (circle)
#1981, Remove unused but set variables as found with gcc 4.6+
#1987, Restore 1.5.x behaviour of ST_Simplify
#1989, Preprocess input geometry to just intersection with raster to be clipped
#1991, geocode really slow on PostgreSQL 9.2
#1996, support POINT EMPTY in GeoJSON output
#1998, Fix ST_{Mod,New}EdgeHeal joining edges sharing both endpoints
#2001, ST_CurveToLine has no effect if the geometry doesn't actually contain an arc
#2015, ST_IsEmpty('POLYGON(EMPTY)') returns False
#2019, ST_FlipCoordinates does not update bbox
#2025, Fix side location conflict at TopoGeo_AddLineString
#2026, improve performance of distance calculations
#2033, Fix adding a splitting point into a 2.5d topology
#2051, Fix excess of precision in ST_AsGeoJSON output
#2052, Fix buffer overflow in lwgeom_to_geojson
#2056, Fixed lack of SRID check of raster and geometry in ST_SetValue()
#2057, Fixed linking issue for raster2psql to libpq
#2060, Fix "dimension" check violation by GetTopoGeomElementArray
#2072, Removed outdated checks preventing ST_Intersects(raster) from working on out-db bands
#2077, Fixed incorrect answers from ST_Hillshade(raster)
#2092, Namespace issue with ST_GeomFromKML,ST_GeomFromGML for libxml 2.8+
#2099, Fix double free on exception in ST_OffsetCurve
#2100, ST_AsRaster() may not return raster with specified pixel type
#2108, Ensure ST_Line_Interpolate_Point always returns POINT
#2109, Ensure ST_Centroid always returns POINT
#2117, Ensure ST_PointOnSurface always returns POINT
#2129, Fix SRID in ST_Homogenize output with collection input
#2130, Fix memory error in MultiPolygon GeoJson parsing
Update URL of Maven jar
#1581, ST_Clip(raster, ...) no longer imposes NODATA on a band if the corresponding band from the source raster did not have NODATA
#1928, Accept array properties in GML input multi-geom input (Kashif Rasul and Shoaib Burq / SpacialDB)
#2082, Add indices on start_node and end_node of topology edge tables
#2087, Speedup topology.GetRingEdges using a recursive CTE
Release date: 2012/06/22
This is a bug fix release, addressing issues that have been filed since the 2.0.0 release.
#1264, fix st_dwithin(geog, geog, 0).
#1468 shp2pgsql-gui table column schema get shifted
#1694, fix building with clang. (vince)
#1708, improve restore of pre-PostGIS 2.0 backups.
#1714, more robust handling of high topology tolerance.
#1755, ST_GeographyFromText support for higher dimensions.
#1759, loading transformed shapefiles in raster enabled db.
#1761, handling of subdatasets in NetCDF, HDF4 and HDF5 in raster2pgsql.
#1763, topology.toTopoGeom use with custom search_path.
#1766, don't let ST_RemEdge* destroy peripheral TopoGeometry objects.
#1774, Clearer error on setting an edge geometry to an invalid one.
#1775, ST_ChangeEdgeGeom collision detection with 2-vertex target.
#1776, fix ST_SymDifference(empty, geom) to return geom.
#1779, install SQL comment files.
#1782, fix spatial reference string handling in raster.
#1789, fix false edge-node crossing report in ValidateTopology.
#1790, fix toTopoGeom handling of duplicated primitives.
#1791, fix ST_Azimuth with very close but distinct points.
#1797, fix (ValidateTopology(xxx)).* syntax calls.
#1805, put back the 900913 SRID entry.
#1813, Only show readable relations in metadata tables.
#1819, fix floating point issues with ST_World2RasterCoord and ST_Raster2WorldCoord variants.
#1820 compilation on 9.2beta1.
#1822, topology load on PostgreSQL 9.2beta1.
#1825, fix prepared geometry cache lookup
#1829, fix uninitialized read in GeoJSON parser
#1834, revise postgis extension to only backup user specified spatial_ref_sys
#1839, handling of subdatasets in GeoTIFF in raster2pgsql.
#1840, fix logic of when to compute # of tiles in raster2pgsql.
#1851, fix spatial_ref_system parameters for EPSG:3844
#1857, fix failure to detect endpoint mismatch in ST_AddEdge*Face*
#1865, data loss in postgis_restore.pl when data rows have leading dashes.
#1867, catch invalid topology name passed to topogeo_add*
#1872, fix ST_ApproxSummarystats to prevent division by zero
#1873, fix ptarray_locate_point to return interpolated Z/M values for on-the-line case
#1875, ST_SummaryStats returns NULL for all parameters except count when count is zero
#1881, shp2pgsql-gui -- editing a field sometimes triggers removing row
#1883, Geocoder install fails trying to run create_census_base_tables() (Brian Panulla)
Release date: 2012/04/03
This is a major release. A hard upgrade is required. Yes this means a full dump reload and some special preparations if you are using obsolete functions. Refer to Section 2.10.2, “Hard upgrade” for details on upgrading. Refer to Section 14.11.6, “PostGIS Functions new, behavior changed, or enhanced in 2.0” for more details and changed/new functions.
We are most indebted to the numerous members in the PostGIS community who were brave enough to test out the new features in this release. No major release can be successful without these folk.
Below are those who have been most valiant, provided very detailed and thorough bug reports, and detailed analysis.
Andrea Peri - Lots of testing on topology, checking for correctness |
Andreas Forø Tollefsen - raster testing |
Chris English - topology stress testing loader functions |
Salvatore Larosa - topology robustness testing |
Brian Hamlin - Benchmarking (also experimental experimental branches before they are folded into core) , general testing of various pieces including Tiger and Topology. Testing on various server VMs |
Mike Pease - Tiger geocoder testing - very detailed reports of issues |
Tom van Tilburg - raster testing |
#722, #302, Most deprecated functions removed (over 250 functions) (Regina Obe, Paul Ramsey)
Unknown SRID changed from -1 to 0. (Paul Ramsey)
-- (most deprecated in 1.2) removed non-ST variants buffer, length, intersects (and internal functions renamed) etc.
-- If you have been using deprecated functions CHANGE your apps or suffer the consequences. If you don't see a function documented -- it ain't supported or it is an internal function. Some constraints in older tables were built with deprecated functions. If you restore you may need to rebuild table constraints with populate_geometry_columns(). If you have applications or tools that rely on deprecated functions, please refer to Q: 3.2 for more details.
#944 geometry_columns is now a view instead of a table (Paul Ramsey, Regina Obe) for tables created the old way reads (srid, type, dims) constraints for geometry columns created with type modifiers reads rom column definition
#1081, #1082, #1084, #1088 - Mangement functions support typmod geometry column creation functions now default to typmod creation (Regina Obe)
#1083 probe_geometry_columns(), rename_geometry_table_constraints(), fix_geometry_columns(); removed - now obsolete with geometry_column view (Regina Obe)
#817 Renaming old 3D functions to the convention ST_3D (Nicklas Avén)
#548 (sorta), ST_NumGeometries,ST_GeometryN now returns 1 (or the geometry) instead of null for single geometries (Sandro Santilli, Maxime van Noppen)
Support for TIN and PolyHedralSurface and enhancement of many functions to support 3D (Olivier Courtin / Oslandia)
Raster support integrated and documented (Pierre Racine, Jorge Arévalo, Mateusz Loskot, Sandro Santilli, David Zwarg, Regina Obe, Bborie Park) (Company developer and funding: University Laval, Deimos Space, CadCorp, Michigan Tech Research Institute, Azavea, Paragon Corporation, UC Davis Center for Vectorborne Diseases)
Making spatial indexes 3D aware - in progress (Paul Ramsey, Mark Cave-Ayland)
Topology support improved (more functions), documented, testing (Sandro Santilli / Faunalia for RT-SIGTA), Andrea Peri, Regina Obe, Jose Carlos Martinez Llari
3D relationship and measurement support functions (Nicklas Avén)
ST_3DDistance, ST_3DClosestPoint, ST_3DIntersects, ST_3DShortestLine and more...
N-Dimensional spatial indexes (Paul Ramsey / OpenGeo)
ST_Split (Sandro Santilli / Faunalia for RT-SIGTA)
ST_IsValidDetail (Sandro Santilli / Faunalia for RT-SIGTA)
ST_MakeValid (Sandro Santilli / Faunalia for RT-SIGTA)
ST_RemoveRepeatedPoints (Sandro Santilli / Faunalia for RT-SIGTA)
ST_GeometryN and ST_NumGeometries support for non-collections (Sandro Santilli)
ST_IsCollection (Sandro Santilli, Maxime van Noppen)
ST_SharedPaths (Sandro Santilli / Faunalia for RT-SIGTA)
ST_Snap (Sandro Santilli)
ST_RelateMatch (Sandro Santilli / Faunalia for RT-SIGTA)
ST_ConcaveHull (Regina Obe and Leo Hsu / Paragon Corporation)
ST_UnaryUnion (Sandro Santilli / Faunalia for RT-SIGTA)
ST_AsX3D (Regina Obe / Arrival 3D funding)
ST_OffsetCurve (Sandro Santilli, Rafal Magda)
ST_GeomFromGeoJSON (Kashif Rasul, Paul Ramsey / Vizzuality funding)
Made shape file loader tolerant of truncated multibyte values found in some free worldwide shapefiles (Sandro Santilli)
Lots of bug fixes and enhancements to shp2pgsql Beefing up regression tests for loaders Reproject support for both geometry and geography during import (Jeff Adams / Azavea, Mark Cave-Ayland)
pgsql2shp conversion from predefined list (Loic Dachary / Mark Cave-Ayland)
Shp-pgsql GUI loader - support loading multiple files at a time. (Mark Leslie)
Extras - upgraded tiger_geocoder from using old TIGER format to use new TIGER shp and file structure format (Stephen Frost)
Extras - revised tiger_geocoder to work with TIGER census 2010 data, addition of reverse geocoder function, various bug fixes, accuracy enhancements, limit max result return, speed improvements, loading routines. (Regina Obe, Leo Hsu / Paragon Corporation / funding provided by Hunter Systems Group)
Overall Documentation proofreading and corrections. (Kasif Rasul)
Cleanup PostGIS JDBC classes, revise to use Maven build. (Maria Arias de Reyna, Sandro Santilli)
#1335 ST_AddPoint returns incorrect result on Linux (Even Rouault)
We thank U.S Department of State Human Information Unit (HIU) and Vizzuality for general monetary support to get PostGIS 2.0 out the door.
Release date: 2012/05/07
This is a bug fix release, addressing issues that have been filed since the 1.5.3 release.
#547, ST_Contains memory problems (Sandro Santilli)
#621, Problem finding intersections with geography (Paul Ramsey)
#627, PostGIS/PostgreSQL process die on invalid geometry (Paul Ramsey)
#810, Increase accuracy of area calculation (Paul Ramsey)
#852, improve spatial predicates robustness (Sandro Santilli, Nicklas Avén)
#877, ST_Estimated_Extent returns NULL on empty tables (Sandro Santilli)
#1028, ST_AsSVG kills whole postgres server when fails (Paul Ramsey)
#1056, Fix boxes of arcs and circle stroking code (Paul Ramsey)
#1121, populate_geometry_columns using deprecated functions (Regin Obe, Paul Ramsey)
#1135, improve testsuite predictability (Andreas 'ads' Scherbaum)
#1146, images generator crashes (bronaugh)
#1170, North Pole intersection fails (Paul Ramsey)
#1179, ST_AsText crash with bad value (kjurka)
#1184, honour DESTDIR in documentation Makefile (Bryce L Nordgren)
#1227, server crash on invalid GML
#1252, SRID appearing in WKT (Paul Ramsey)
#1264, st_dwithin(g, g, 0) doesn't work (Paul Ramsey)
#1344, allow exporting tables with invalid geometries (Sandro Santilli)
#1389, wrong proj4text for SRID 31300 and 31370 (Paul Ramsey)
#1406, shp2pgsql crashes when loading into geography (Sandro Santilli)
#1595, fixed SRID redundancy in ST_Line_SubString (Sandro Santilli)
#1596, check SRID in UpdateGeometrySRID (Mike Toews, Sandro Santilli)
#1602, fix ST_Polygonize to retain Z (Sandro Santilli)
#1697, fix crash with EMPTY entries in GiST index (Paul Ramsey)
#1772, fix ST_Line_Locate_Point with collapsed input (Sandro Santilli)
#1799, Protect ST_Segmentize from max_length=0 (Sandro Santilli)
Alter parameter order in 900913 (Paul Ramsey)
Support builds with "gmake" (Greg Troxel)
Release date: 2011/06/25
This is a bug fix release, addressing issues that have been filed since the 1.5.2 release. If you are running PostGIS 1.3+, a soft upgrade is sufficient otherwise a hard upgrade is recommended.
#1056, produce correct bboxes for arc geometries, fixes index errors (Paul Ramsey)
#1007, ST_IsValid crash fix requires GEOS 3.3.0+ or 3.2.3+ (Sandro Santilli, reported by Birgit Laggner)
#940, support for PostgreSQL 9.1 beta 1 (Regina Obe, Paul Ramsey, patch submitted by stl)
#845, ST_Intersects precision error (Sandro Santilli, Nicklas Avén) Reported by cdestigter
#884, Unstable results with ST_Within, ST_Intersects (Chris Hodgson)
#779, shp2pgsql -S option seems to fail on points (Jeff Adams)
#666, ST_DumpPoints is not null safe (Regina Obe)
#631, Update NZ projections for grid transformation support (jpalmer)
#630, Peculiar Null treatment in arrays in ST_Collect (Chris Hodgson) Reported by David Bitner
#624, Memory leak in ST_GeogFromText (ryang, Paul Ramsey)
#609, Bad source code in manual section 5.2 Java Clients (simoc, Regina Obe)
#604, shp2pgsql usage touchups (Mike Toews, Paul Ramsey)
#573 ST_Union fails on a group of linestrings Not a PostGIS bug, fixed in GEOS 3.3.0
#457 ST_CollectionExtract returns non-requested type (Nicklas Avén, Paul Ramsey)
#441 ST_AsGeoJson Bbox on GeometryCollection error (Olivier Courtin)
#411 Ability to backup invalid geometries (Sando Santilli) Reported by Regione Toscana
#409 ST_AsSVG - degraded (Olivier Courtin) Reported by Sdikiy
#373 Documentation syntax error in hard upgrade (Paul Ramsey) Reported by psvensso
Release date: 2010/09/27
This is a bug fix release, addressing issues that have been filed since the 1.5.1 release. If you are running PostGIS 1.3+, a soft upgrade is sufficient otherwise a hard upgrade is recommended.
Loader: fix handling of empty (0-verticed) geometries in shapefiles. (Sandro Santilli)
#536, Geography ST_Intersects, ST_Covers, ST_CoveredBy and Geometry ST_Equals not using spatial index (Regina Obe, Nicklas Aven)
#573, Improvement to ST_Contains geography (Paul Ramsey)
Loader: Add support for command-q shutdown in Mac GTK build (Paul Ramsey)
#393, Loader: Add temporary patch for large DBF files (Maxime Guillaud, Paul Ramsey)
#507, Fix wrong OGC URN in GeoJSON and GML output (Olivier Courtin)
spatial_ref_sys.sql Add datum conversion for projection SRID 3021 (Paul Ramsey)
Geography - remove crash for case when all geographies are out of the estimate (Paul Ramsey)
#469, Fix for array_aggregation error (Greg Stark, Paul Ramsey)
#532, Temporary geography tables showing up in other user sessions (Paul Ramsey)
#562, ST_Dwithin errors for large geographies (Paul Ramsey)
#513, shape loading GUI tries to make spatial index when loading DBF only mode (Paul Ramsey)
#527, shape loading GUI should always append log messages (Mark Cave-Ayland)
#504, shp2pgsql should rename xmin/xmax fields (Sandro Santilli)
#458, postgis_comments being installed in contrib instead of version folder (Mark Cave-Ayland)
#474, Analyzing a table with geography column crashes server (Paul Ramsey)
#581, LWGEOM-expand produces inconsistent results (Mark Cave-Ayland)
#513, Add dbf filter to shp2pgsql-gui and allow uploading dbf only (Paul Ramsey)
Fix further build issues against PostgreSQL 9.0 (Mark Cave-Ayland)
#572, Password whitespace for Shape File (Mark Cave-Ayland)
#603, shp2pgsql: "-w" produces invalid WKT for MULTI* objects. (Mark Cave-Ayland)
Release date: 2010/03/11
This is a bug fix release, addressing issues that have been filed since the 1.4.1 release. If you are running PostGIS 1.3+, a soft upgrade is sufficient otherwise a hard upgrade is recommended.
#410, update embedded bbox when applying ST_SetPoint, ST_AddPoint ST_RemovePoint to a linestring (Paul Ramsey)
#411, allow dumping tables with invalid geometries (Sandro Santilli, for Regione Toscana-SIGTA)
#414, include geography_columns view when running upgrade scripts (Paul Ramsey)
#419, allow support for multilinestring in ST_Line_Substring (Paul Ramsey, for Lidwala Consulting Engineers)
#421, fix computed string length in ST_AsGML() (Olivier Courtin)
#441, fix GML generation with heterogeneous collections (Olivier Courtin)
#443, incorrect coordinate reversal in GML 3 generation (Olivier Courtin)
#450, #451, wrong area calculation for geography features that cross the date line (Paul Ramsey)
Ensure support for upcoming 9.0 PgSQL release (Paul Ramsey)
Release date: 2010/02/04
This release provides support for geographic coordinates (lat/lon) via a new GEOGRAPHY type. Also performance enhancements, new input format support (GML,KML) and general upkeep.
The public API of PostGIS will not change during minor (0.0.X) releases.
The definition of the =~ operator has changed from an exact geometric equality check to a bounding box equality check.
GEOS, Proj4, and LibXML2 are now mandatory dependencies
The library versions below are the minimum requirements for PostGIS 1.5
PostgreSQL 8.3 and higher on all platforms
GEOS 3.1 and higher only (GEOS 3.2+ to take advantage of all features)
LibXML2 2.5+ related to new ST_GeomFromGML/KML functionality
Proj4 4.5 and higher only
Section 14.11.8, “PostGIS Functions new, behavior changed, or enhanced in 1.5”
Added Hausdorff distance calculations (#209) (Vincent Picavet)
Added parameters argument to ST_Buffer operation to support one-sided buffering and other buffering styles (Sandro Santilli)
Addition of other Distance related visualization and analysis functions (Nicklas Aven)
ST_ClosestPoint
ST_DFullyWithin
ST_LongestLine
ST_MaxDistance
ST_ShortestLine
ST_DumpPoints (Maxime van Noppen)
KML, GML input via ST_GeomFromGML and ST_GeomFromKML (Olivier Courtin)
Extract homogeneous collection with ST_CollectionExtract (Paul Ramsey)
Add measure values to an existing linestring with ST_AddMeasure (Paul Ramsey)
History table implementation in utils (George Silva)
Geography type and supporting functions
Spherical algorithms (Dave Skea)
Object/index implementation (Paul Ramsey)
Selectivity implementation (Mark Cave-Ayland)
Serializations to KML, GML and JSON (Olivier Courtin)
ST_Area, ST_Distance, ST_DWithin, ST_GeogFromText, ST_GeogFromWKB, ST_Intersects, ST_Covers, ST_Buffer (Paul Ramsey)
Performance improvements to ST_Distance (Nicklas Aven)
Documentation updates and improvements (Regina Obe, Kevin Neufeld)
Testing and quality control (Regina Obe)
PostGIS 1.5 support PostgreSQL 8.5 trunk (Guillaume Lelarge)
Win32 support and improvement of core shp2pgsql-gui (Mark Cave-Ayland)
In place 'make check' support (Paul Ramsey)
Release date: 2009/07/24
This release provides performance enhancements, improved internal structures and testing, new features, and upgraded documentation. If you are running PostGIS 1.1+, a soft upgrade is sufficient otherwise a hard upgrade is recommended.
As of the 1.4 release series, the public API of PostGIS will not change during minor releases.
The versions below are the *minimum* requirements for PostGIS 1.4
PostgreSQL 8.2 and higher on all platforms
GEOS 3.0 and higher only
PROJ4 4.5 and higher only
ST_Union() uses high-speed cascaded union when compiled against GEOS 3.1+ (Paul Ramsey)
ST_ContainsProperly() requires GEOS 3.1+
ST_Intersects(), ST_Contains(), ST_Within() use high-speed cached prepared geometry against GEOS 3.1+ (Paul Ramsey / funded by Zonar Systems)
Vastly improved documentation and reference manual (Regina Obe & Kevin Neufeld)
Figures and diagram examples in the reference manual (Kevin Neufeld)
ST_IsValidReason() returns readable explanations for validity failures (Paul Ramsey)
ST_GeoHash() returns a geohash.org signature for geometries (Paul Ramsey)
GTK+ multi-platform GUI for shape file loading (Paul Ramsey)
ST_LineCrossingDirection() returns crossing directions (Paul Ramsey)
ST_LocateBetweenElevations() returns sub-string based on Z-ordinate. (Paul Ramsey)
Geometry parser returns explicit error message about location of syntax errors (Mark Cave-Ayland)
ST_AsGeoJSON() return JSON formatted geometry (Olivier Courtin)
Populate_Geometry_Columns() -- automatically add records to geometry_columns for TABLES and VIEWS (Kevin Neufeld)
ST_MinimumBoundingCircle() -- returns the smallest circle polygon that can encompass a geometry (Bruce Rindahl)
Core geometry system moved into independent library, liblwgeom. (Mark Cave-Ayland)
New build system uses PostgreSQL "pgxs" build bootstrapper. (Mark Cave-Ayland)
Debugging framework formalized and simplified. (Mark Cave-Ayland)
All build-time #defines generated at configure time and placed in headers for easier cross-platform support (Mark Cave-Ayland)
Logging framework formalized and simplified (Mark Cave-Ayland)
Expanded and more stable support for CIRCULARSTRING, COMPOUNDCURVE and CURVEPOLYGON, better parsing, wider support in functions (Mark Leslie & Mark Cave-Ayland)
Improved support for OpenSolaris builds (Paul Ramsey)
Improved support for MSVC builds (Mateusz Loskot)
Updated KML support (Olivier Courtin)
Unit testing framework for liblwgeom (Paul Ramsey)
New testing framework to comprehensively exercise every PostGIS function (Regine Obe)
Performance improvements to all geometry aggregate functions (Paul Ramsey)
Support for the upcoming PostgreSQL 8.4 (Mark Cave-Ayland, Talha Bin Rizwan)
Shp2pgsql and pgsql2shp re-worked to depend on the common parsing/unparsing code in liblwgeom (Mark Cave-Ayland)
Use of PDF DbLatex to build PDF docs and preliminary instructions for build (Jean David Techer)
Automated User documentation build (PDF and HTML) and Developer Doxygen Documentation (Kevin Neufeld)
Automated build of document images using ImageMagick from WKT geometry text files (Kevin Neufeld)
More attractive CSS for HTML documentation (Dane Springmeyer)
Release date: 2009/05/04
If you are running PostGIS 1.1+, a soft upgrade is sufficient otherwise a hard upgrade is recommended. This release adds support for PostgreSQL 8.4, exporting prj files from the database with shape data, some crash fixes for shp2pgsql, and several small bug fixes in the handling of "curve" types, logical error importing dbf only files, improved error handling of AddGeometryColumns.
Release date: 2008/12/15
If you are running PostGIS 1.1+, a soft upgrade is sufficient otherwise a hard upgrade is recommended. This release is a bug fix release to address a failure in ST_Force_Collection and related functions that critically affects using MapServer with LINE layers.
Release date: 2008/11/24
This release adds support for GeoJSON output, building with PostgreSQL 8.4, improves documentation quality and output aesthetics, adds function-level SQL documentation, and improves performance for some spatial predicates (point-in-polygon tests).
Bug fixes include removal of crashers in handling circular strings for many functions, some memory leaks removed, a linear referencing failure for measures on vertices, and more. See the NEWS file for details.
Release date: 2008/04/12
This release fixes bugs shp2pgsql, adds enhancements to SVG and KML support, adds a ST_SimplifyPreserveTopology function, makes the build more sensitive to GEOS versions, and fixes a handful of severe but rare failure cases.
Release date: 2007/12/01
This release fixes bugs in ST_EndPoint() and ST_Envelope, improves support for JDBC building and OS/X, and adds better support for GML output with ST_AsGML(), including GML3 output.
Release date: 2007/08/13
This release fixes some oversights in the previous release around version numbering, documentation, and tagging.
Release date: 2007/08/09
This release provides performance enhancements to the relational functions, adds new relational functions and begins the migration of our function names to the SQL-MM convention, using the spatial type (SP) prefix.
JDBC: Added Hibernate Dialect (thanks to Norman Barker)
Added ST_Covers and ST_CoveredBy relational functions. Description and justification of these functions can be found at http://lin-ear-th-inking.blogspot.com/2007/06/subtleties-of-ogc-covers-spatial.html
Added ST_DWithin relational function.
Added cached and indexed point-in-polygon short-circuits for the functions ST_Contains, ST_Intersects, ST_Within and ST_Disjoint
Added inline index support for relational functions (except ST_Disjoint)
Release date: 2007/01/11
This release provides bug fixes in PostgreSQL 8.2 support and some small performance enhancements.
Fixed point-in-polygon shortcut bug in Within().
Fixed PostgreSQL 8.2 NULL handling for indexes.
Updated RPM spec files.
Added short-circuit for Transform() in no-op case.
JDBC: Fixed JTS handling for multi-dimensional geometries (thanks to Thomas Marti for hint and partial patch). Additionally, now JavaDoc is compiled and packaged. Fixed classpath problems with GCJ. Fixed pgjdbc 8.2 compatibility, losing support for jdk 1.3 and older.
Release date: 2006/12/08
This release provides type definitions along with serialization/deserialization capabilities for SQL-MM defined curved geometries, as well as performance enhancements.
Release date: 2006/11/02
This is a bugfix release, in particular fixing a critical error with GEOS interface in 64bit systems. Includes an updated of the SRS parameters and an improvement in reprojections (take Z in consideration). Upgrade is encouraged.
If you are upgrading from release 1.0.3 or later follow the soft upgrade procedure.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
fixed CAPI change that broke 64-bit platforms
loader/dumper: fixed regression tests and usage output
Fixed setSRID() bug in JDBC, thanks to Thomas Marti
use Z ordinate in reprojections
spatial_ref_sys.sql updated to EPSG 6.11.1
Simplified Version.config infrastructure to use a single pack of version variables for everything.
Include the Version.config in loader/dumper USAGE messages
Replace hand-made, fragile JDBC version parser with Properties
Release date: 2006/10/13
This is an bugfix release, including a critical segfault on win32. Upgrade is encouraged.
If you are upgrading from release 1.0.3 or later follow the soft upgrade procedure.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
Fixed MingW link error that was causing pgsql2shp to segfault on Win32 when compiled for PostgreSQL 8.2
fixed nullpointer Exception in Geometry.equals() method in Java
Added EJB3Spatial.odt to fulfill the GPL requirement of distributing the "preferred form of modification"
Removed obsolete synchronization from JDBC Jts code.
Updated heavily outdated README files for shp2pgsql/pgsql2shp by merging them with the manpages.
Fixed version tag in jdbc code that still said "1.1.3" in the "1.1.4" release.
Release date: 2006/09/27
This is an bugfix release including some improvements in the Java interface. Upgrade is encouraged.
If you are upgrading from release 1.0.3 or later follow the soft upgrade procedure.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
Fixed support for PostgreSQL 8.2
Fixed bug in collect() function discarding SRID of input
Added SRID match check in MakeBox2d and MakeBox3d
Fixed regress tests to pass with GEOS-3.0.0
Improved pgsql2shp run concurrency.
reworked JTS support to reflect new upstream JTS developers' attitude to SRID handling. Simplifies code and drops build depend on GNU trove.
Added EJB2 support generously donated by the "Geodetix s.r.l. Company"
Added EJB3 tutorial / examples donated by Norman Barker <nbarker@ittvis.com>
Reorganized java directory layout a little.
Release date: 2006/06/30
This is an bugfix release including also some new functionalities (most notably long transaction support) and portability enhancements. Upgrade is encouraged.
If you are upgrading from release 1.0.3 or later follow the soft upgrade procedure.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
BUGFIX in distance(poly,poly) giving wrong results.
BUGFIX in pgsql2shp successful return code.
BUGFIX in shp2pgsql handling of MultiLine WKT.
BUGFIX in affine() failing to update bounding box.
WKT parser: forbidden construction of multigeometries with EMPTY elements (still supported for GEOMETRYCOLLECTION).
NEW Long Transactions support.
NEW DumpRings() function.
NEW AsHEXEWKB(geom, XDR|NDR) function.
Improved regression tests: MultiPoint and scientific ordinates
Fixed some minor bugs in jdbc code
Added proper accessor functions for all fields in preparation of making those fields private later
Release date: 2006/03/30
This is an bugfix release including some new functions and portability enhancements. Upgrade is encouraged.
If you are upgrading from release 1.0.3 or later follow the soft upgrade procedure.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
BUGFIX in SnapToGrid() computation of output bounding box
BUGFIX in EnforceRHR()
jdbc2 SRID handling fixes in JTS code
Fixed support for 64bit archs
Regress tests can now be run *before* postgis installation
New affine() matrix transformation functions
New rotate{,X,Y,Z}() function
Old translating and scaling functions now use affine() internally
Embedded access control in estimated_extent() for builds against pgsql >= 8.0.0
Release date: 2006/01/23
This is an important Bugfix release, upgrade is highly recommended. Previous version contained a bug in postgis_restore.pl preventing hard upgrade procedure to complete and a bug in GEOS-2.2+ connector preventing GeometryCollection objects to be used in topological operations.
If you are upgrading from release 1.0.3 or later follow the soft upgrade procedure.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
Fixed a premature exit in postgis_restore.pl
BUGFIX in geometrycollection handling of GEOS-CAPI connector
Solaris 2.7 and MingW support improvements
BUGFIX in line_locate_point()
Fixed handling of postgresql paths
BUGFIX in line_substring()
Added support for localized cluster in regress tester
Release date: 2005/12/21
This is a Minor release, containing many improvements and new things. Most notably: build procedure greatly simplified; transform() performance drastically improved; more stable GEOS connectivity (CAPI support); lots of new functions; draft topology support.
It is highly recommended that you upgrade to GEOS-2.2.x before installing PostGIS, this will ensure future GEOS upgrades won't require a rebuild of the PostGIS library.
This release includes code from Mark Cave Ayland for caching of proj4 objects. Markus Schaber added many improvements in his JDBC2 code. Alex Bodnaru helped with PostgreSQL source dependency relief and provided Debian specfiles. Michael Fuhr tested new things on Solaris arch. David Techer and Gerald Fenoy helped testing GEOS C-API connector. Hartmut Tschauner provided code for the azimuth() function. Devrim GUNDUZ provided RPM specfiles. Carl Anderson helped with the new area building functions. See the credits section for more names.
If you are upgrading from release 1.0.3 or later you DO NOT need a dump/reload. Simply sourcing the new lwpostgis_upgrade.sql script in all your existing databases will work. See the soft upgrade chapter for more information.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
scale() and transscale() companion methods to translate()
line_substring()
line_locate_point()
M(point)
LineMerge(geometry)
shift_longitude(geometry)
azimuth(geometry)
locate_along_measure(geometry, float8)
locate_between_measures(geometry, float8, float8)
SnapToGrid by point offset (up to 4d support)
BuildArea(any_geometry)
OGC BdPolyFromText(linestring_wkt, srid)
OGC BdMPolyFromText(linestring_wkt, srid)
RemovePoint(linestring, offset)
ReplacePoint(linestring, offset, point)
Fixed memory leak in polygonize()
Fixed bug in lwgeom_as_anytype cast functions
Fixed USE_GEOS, USE_PROJ and USE_STATS elements of postgis_version() output to always reflect library state.
SnapToGrid doesn't discard higher dimensions
Changed Z() function to return NULL if requested dimension is not available
Much faster transform() function, caching proj4 objects
Removed automatic call to fix_geometry_columns() in AddGeometryColumns() and update_geometry_stats()
Makefile improvements
JTS support improvements
Improved regression test system
Basic consistency check method for geometry collections
Support for (Hex)(E)wkb
Autoprobing DriverWrapper for HexWKB / EWKT switching
fix compile problems in ValueSetter for ancient jdk releases.
fix EWKT constructors to accept SRID=4711; representation
added preliminary read-only support for java2d geometries
Full autoconf-based configuration, with PostgreSQL source dependency relief
GEOS C-API support (2.2.0 and higher)
Initial support for topology modelling
Debian and RPM specfiles
New lwpostgis_upgrade.sql script
Release date: 2005/12/06
Contains a few bug fixes and improvements.
If you are upgrading from release 1.0.3 or later you DO NOT need a dump/reload.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
Fixed palloc(0) call in collection deserializer (only gives problem with --enable-cassert)
Fixed bbox cache handling bugs
Fixed geom_accum(NULL, NULL) segfault
Fixed segfault in addPoint()
Fixed short-allocation in lwcollection_clone()
Fixed bug in segmentize()
Fixed bbox computation of SnapToGrid output
Release date: 2005/11/25
Contains memory-alignment fixes in the library, a segfault fix in loader's handling of UTF8 attributes and a few improvements and cleanups.
Return code of shp2pgsql changed from previous releases to conform to unix standards (return 0 on success). |
If you are upgrading from release 1.0.3 or later you DO NOT need a dump/reload.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
Fixed memory alignment problems
Fixed computation of null values fraction in analyzer
Fixed a small bug in the getPoint4d_p() low-level function
Speedup of serializer functions
Fixed a bug in force_3dm(), force_3dz() and force_4d()
Fixed return code of shp2pgsql
Fixed back-compatibility issue in loader (load of null shapefiles)
Fixed handling of trailing dots in dbf numerical attributes
Segfault fix in shp2pgsql (utf8 encoding)
Release date: 2005/09/09
Contains important bug fixes and a few improvements. In particular, it fixes a memory leak preventing successful build of GiST indexes for large spatial tables.
If you are upgrading from release 1.0.3 you DO NOT need a dump/reload.
If you are upgrading from a release between 1.0.0RC6 and 1.0.2 (inclusive) and really want a live upgrade read the upgrade section of the 1.0.3 release notes chapter.
Upgrade from any release prior to 1.0.0RC6 requires an hard upgrade.
Memory leak plugged in GiST indexing
Segfault fix in transform() handling of proj4 errors
Fixed some proj4 texts in spatial_ref_sys (missing +proj)
Loader: fixed string functions usage, reworked NULL objects check, fixed segfault on MULTILINESTRING input.
Fixed bug in MakeLine dimension handling
Fixed bug in translate() corrupting output bounding box
Release date: 2005/08/08
Contains some bug fixes - including a severe one affecting correctness of stored geometries - and a few improvements.
Due to a bug in a bounding box computation routine, the upgrade procedure requires special attention, as bounding boxes cached in the database could be incorrect.
An hard upgrade procedure (dump/reload) will force recomputation of all bounding boxes (not included in dumps). This is required if upgrading from releases prior to 1.0.0RC6.
If you are upgrading from versions 1.0.0RC6 or up, this release includes a perl script (utils/rebuild_bbox_caches.pl) to force recomputation of geometries' bounding boxes and invoke all operations required to propagate eventual changes in them (geometry statistics update, reindexing). Invoke the script after a make install (run with no args for syntax help). Optionally run utils/postgis_proc_upgrade.pl to refresh postgis procedures and functions signatures (see Soft upgrade).
Severe bugfix in lwgeom's 2d bounding box computation
Bugfix in WKT (-w) POINT handling in loader
Bugfix in dumper on 64bit machines
Bugfix in dumper handling of user-defined queries
Bugfix in create_undef.pl script
Release date: 2005/07/04
Contains a few bug fixes and improvements.
If you are upgrading from release 1.0.0RC6 or up you DO NOT need a dump/reload.
Upgrading from older releases requires a dump/reload. See the upgrading chapter for more informations.
Fault tolerant btree ops
Memory leak plugged in pg_error
Rtree index fix
Cleaner build scripts (avoided mix of CFLAGS and CXXFLAGS)
Release date: 2005/05/24
Contains a few bug fixes and some improvements.
If you are upgrading from release 1.0.0RC6 or up you DO NOT need a dump/reload.
Upgrading from older releases requires a dump/reload. See the upgrading chapter for more informations.
BUGFIX in shp2pgsql escape functions
better support for concurrent postgis in multiple schemas
documentation fixes
jdbc2: compile with "-target 1.2 -source 1.2" by default
NEW -k switch for pgsql2shp
NEW support for custom createdb options in postgis_restore.pl
BUGFIX in pgsql2shp attribute names unicity enforcement
BUGFIX in Paris projections definitions
postgis_restore.pl cleanups
Release date: 2005/04/19
Final 1.0.0 release. Contains a few bug fixes, some improvements in the loader (most notably support for older postgis versions), and more docs.
If you are upgrading from release 1.0.0RC6 you DO NOT need a dump/reload.
Upgrading from any other precedent release requires a dump/reload. See the upgrading chapter for more informations.
BUGFIX in transform() releasing random memory address
BUGFIX in force_3dm() allocating less memory then required
BUGFIX in join selectivity estimator (defaults, leaks, tuplecount, sd)
BUGFIX in shp2pgsql escape of values starting with tab or single-quote
NEW manual pages for loader/dumper
NEW shp2pgsql support for old (HWGEOM) postgis versions
NEW -p (prepare) flag for shp2pgsql
NEW manual chapter about OGC compliancy enforcement
NEW autoconf support for JTS lib
BUGFIX in estimator testers (support for LWGEOM and schema parsing)
Release date: 2005/03/30
Sixth release candidate for 1.0.0. Contains a few bug fixes and cleanups.
You need a dump/reload to upgrade from precedent releases. See the upgrading chapter for more informations.
Release date: 2005/03/25
Fifth release candidate for 1.0.0. Contains a few bug fixes and a improvements.
If you are upgrading from release 1.0.0RC4 you DO NOT need a dump/reload.
Upgrading from any other precedent release requires a dump/reload. See the upgrading chapter for more informations.
BUGFIX (segfaulting) in box3d computation (yes, another!).
BUGFIX (segfaulting) in estimated_extent().
Release date: 2005/03/18
Fourth release candidate for 1.0.0. Contains bug fixes and a few improvements.
You need a dump/reload to upgrade from precedent releases. See the upgrading chapter for more informations.
BUGFIX (segfaulting) in geom_accum().
BUGFIX in 64bit architectures support.
BUGFIX in box3d computation function with collections.
NEW subselects support in selectivity estimator.
Early return from force_collection.
Consistency check fix in SnapToGrid().
Box2d output changed back to 15 significant digits.
NEW distance_sphere() function.
Changed get_proj4_from_srid implementation to use PL/PGSQL instead of SQL.
BUGFIX in loader and dumper handling of MultiLine shapes
BUGFIX in loader, skipping all but first hole of polygons.
jdbc2: code cleanups, Makefile improvements
FLEX and YACC variables set *after* pgsql Makefile.global is included and only if the pgsql *stripped* version evaluates to the empty string
Added already generated parser in release
Build scripts refinements
improved version handling, central Version.config
improvements in postgis_restore.pl
Release date: 2005/02/24
Third release candidate for 1.0.0. Contains many bug fixes and improvements.
You need a dump/reload to upgrade from precedent releases. See the upgrading chapter for more informations.
BUGFIX in transform(): missing SRID, better error handling.
BUGFIX in memory alignment handling
BUGFIX in force_collection() causing mapserver connector failures on simple (single) geometry types.
BUGFIX in GeometryFromText() missing to add a bbox cache.
reduced precision of box2d output.
prefixed DEBUG macros with PGIS_ to avoid clash with pgsql one
plugged a leak in GEOS2POSTGIS converter
Reduced memory usage by early releasing query-context palloced one.
BUGFIX in 72 index bindings.
BUGFIX in probe_geometry_columns() to work with PG72 and support multiple geometry columns in a single table
NEW bool::text cast
Some functions made IMMUTABLE from STABLE, for performance improvement.
jdbc2: small patches, box2d/3d tests, revised docs and license.
jdbc2: bug fix and testcase in for pgjdbc 8.0 type autoregistration
jdbc2: Removed use of jdk1.4 only features to enable build with older jdk releases.
jdbc2: Added support for building against pg72jdbc2.jar
jdbc2: updated and cleaned makefile
jdbc2: added BETA support for jts geometry classes
jdbc2: Skip known-to-fail tests against older PostGIS servers.
jdbc2: Fixed handling of measured geometries in EWKT.
new performance tips chapter in manual
documentation updates: pgsql72 requirement, lwpostgis.sql
few changes in autoconf
BUILDDATE extraction made more portable
fixed spatial_ref_sys.sql to avoid vacuuming the whole database.
spatial_ref_sys: changed Paris entries to match the ones distributed with 0.x.
Release date: 2005/01/26
Second release candidate for 1.0.0 containing bug fixes and a few improvements.
You need a dump/reload to upgrade from precedent releases. See the upgrading chapter for more informations.
BUGFIX in pointarray box3d computation
BUGFIX in distance_spheroid definition
BUGFIX in transform() missing to update bbox cache
NEW jdbc driver (jdbc2)
GEOMETRYCOLLECTION(EMPTY) syntax support for backward compatibility
Faster binary outputs
Stricter OGC WKB/WKT constructors
More correct STABLE, IMMUTABLE, STRICT uses in lwpostgis.sql
stricter OGC WKB/WKT constructors
Release date: 2005/01/13
This is the first candidate of a major postgis release, with internal storage of postgis types redesigned to be smaller and faster on indexed queries.
You need a dump/reload to upgrade from precedent releases. See the upgrading chapter for more informations.
Faster canonical input parsing.
Lossless canonical output.
EWKB Canonical binary IO with PG>73.
Support for up to 4d coordinates, providing lossless shapefile->postgis->shapefile conversion.
New function: UpdateGeometrySRID(), AsGML(), SnapToGrid(), ForceRHR(), estimated_extent(), accum().
Vertical positioning indexed operators.
JOIN selectivity function.
More geometry constructors / editors.
PostGIS extension API.
UTF8 support in loader.