Version 54 (modified by jbronn, 12 years ago) (diff)

updated api, fixed some typgraphical errors


The gis branch intends to add a contrib app allowing for Geographic-enabled fields and queries.


Note: The content herein is a loosely structured collection of notes and links that we have found useful, not necessarily what will be supported in the future.

What's GIS?

  • Series of blog posts giving intro to GIS; choice quote from an early post: "If you feel like ending a conversation with a developer then simply bring up the topic of character encodings ... [o]r ... coordinate systems. ... So in the spirit of Tim Bray's and Joel Spolsky's wonderful writeups of character encodings, I thought I'd put together a basic survival guide to coordinate systems over my next few posts and then tie it back to Google Maps."
  • More on map projections, including why people can't agree on just one (utf-8).
  • geodesy the field of science for this stuff.

Useful Code

  • PostGIS, the OpenGIS SQL Types (pdf) implementation for Postgresql
  • GEOS, low-level C++ port of Jave Topology Suite, used by PostGIS
  • GeoTypes is a type (and conversion) library for PostGIS via psycopg.
  • Geopy
    • Calculates distances using (very accurate) Vincenty, and uses the WGS 84 datum by default.
    • Has utility functions for unit of measurement (UOM) conversions (e.g. meters -> kilometers, kilometers -> miles, etc.)
    • Excellent GeoCoding capabilites. Has interfaces for Google, Yahoo, Microsoft Live, MediaWiki, and
  • GDAL/OGR, a library for fiddling with raster geo images.
    • Has a Python interface. A SWIG interface is in development, but not yet stable (no access to full API).
    • shapelib and ogr2ogr are useful for ESRI shapefile manipulations. ESRI shapefiles are a lingua frana GIS format.
  • Geo::Coder::US An excellent Perl library for GeoCoding that powers Users can create their own Geographic databases using the Census Bureau's TIGER/Line data (see below).
  • GeoRosetta, CC-BY-SA licensed, quality-controlled, collection of geocoding data. Not yet released to public(?).
  • MapServer: University of Minnesota (UMN) "open source development environment for building spatially-enabled internet applications."
  • MapNik: C++ and Python toolkit for developing mapping applications. Claimed benefits over MapServer: "It uses the AGG library and offers world class anti-aliasing rendering with subpixel accuracy for geographic data. It is written from scratch in modern C++ and doesn't suffer from design decisions made a decade ago." See MapNik FAQ.
  • Ruby on Rails
    • IvyGIS: Google-maps type displays with RoR and UMN's MapServer
    • Spatial Adapter for Rails: A plugin for Rails which manages the MySql Spatial and PostGIS geometric columns in a transparent way (that is like the other base data type columns). This might have some useful techniques for when we try to support other spatial extensions other than PostGIS.
    • Cartographer GMaps plugin

Useful Data

  • TIGER/Line: "The TIGER/Line files are extracts of selected geographic and cartographic information from the Census Bureau's TIGER® (Topologically Integrated Geographic Encoding and Referencing) database." This data is useful in creating your own geocoding database service. Currently 2006 First Edition is the latest, but second edition should be coming soon. Note: The Census Bureau will be providing SHP files in Fall, 2007.


  • Place your questions here.
  • Q: When dealing with points (say, degrees) from, do they need to be converted to be useful on the back-end data, assuming -that- data is in degrees? Is it enough to have the same datum and origin? (Reading the intro above is likely to answer the question.)
    • My (JDunck) reading indicates yes. Given the same coordinate system (i.e. datum, origin, and axes), degrees are useful without conversion.
  • Q: Can this implementation work with MySQL spatial-extensions. If not, it's planned?
    • No. Yes it's (now) planned, see Phase 3 below. From the last time I (jbronn) checked, MySQL's spatial capabilities have improved. However, we're going to focus our efforts on PostGIS until things are worked out a bit more -- as a spatial database it is more standards compliant (OpenGIS consortium), more widely used, and has more features (e.g. coordinate transformation, geometry_columns and spatial_ref_sys tables). It is definitely something I would want to implement in the future since I do like MySQL.


Phase 1

  • Create Geometry-enabled fields and manager. Status: complete as of r4788.
  • Allow for Geometry-enabled queries. Status: complete as of r4788.

Phase 2

  • Add as much from the PostGIS API as possible.
  • Finish PostGIS indexing capability (complete).
  • Admin fields and forms.
  • Add geometry-enabled routines to the fields that call directly on GEOS routines -- like area(), centroid(), etc.
  • Support for mapping frameworks.
  • Utilities for importing raster data (SHP files first) directly into Django models.

Phase 3

  • Support MySQL databases.


  • PCL (Python Cartographic Library), now part of GIS Python, has done a lot of good work already. Let's apply the DRY principle.
  • Strong opportunities for collaboration with regards to:
    • Mapping framework
    • Utilities
    • Database representation ideas
    • WMS/WMF Framework
    • GEOS support, Sean Gilles (lead developer of PCL) looking for help maintaining Python/SWIG interface to GEOS. If SWIG interface no longer maintained, might have to move to PCL for up-to-date GEOS library support.


Here is an example of how the model API currently works (assume this example is in geo_app/

from django.contrib.gis.db import models

class District(models.Model, models.GeoMixin):
    name = models.CharField(maxlength=35)
    num  = models.IntegerField()
    poly = models.PolygonField()

    objects = models.GeoManager()

class School(models.Model, models.GeoMixin):
    name  = models.CharField(maxlength=35)
    point = models.PointField(index=True)

    objects = models.GeoManager()

Notes: The GeoMixin class allows for extra instance methods. The index keyword is used to indicate that a GiST index be created for the School PointFields fields.

Use the just like you normally would:

$ python sqlall geo_app
CREATE TABLE "geo_app_school" (
    "id" serial NOT NULL PRIMARY KEY,
    "name" varchar(35) NOT NULL
CREATE TABLE "geo_app_district" (
    "id" serial NOT NULL PRIMARY KEY,
    "name" varchar(35) NOT NULL,
    "num" integer NOT NULL
SELECT AddGeometryColumn('geo_app_school', 'point', 4326, 'POINT', 2);
CREATE INDEX "geo_app_school_point_id" ON "geo_app_school" USING GIST ( "point" GIST_GEOMETRY_OPS );
SELECT AddGeometryColumn('geo_app_district', 'poly', 4326, 'MULTIPOLYGON', 2);
$ python syncdb geo_app

PostGIS additions to the API may now be used. Geographic queries are done by calling geo_filter() and geo_exclude on geometry-enabled models. In the following example, the bbcontains lookup type is used which is the same as the PostGIS && operator. It looks to see if the bounding box of the polygon contains the specific point. The next example uses the PostGIS Contains() function, which calls GEOS library to test if the polygon actually contains the specific point, not just the bounding box.

>>> from geo_app.models import District, School
>>> qs1 = District.objects.geo_filter(poly__bbcontains='POINT(-95.362293 29.756539)') 
>>> qs2 = District.objects.geo_filter(poly__contains='POINT(-95.362293 29.756539)') 

Both geo_filter() and filter() may be used in the same query. For example, the following query set will only show school districts that have 'Houston' in their name and contain the given point within their polygon boundary:

>>> qs = District.objects.filter(name__contains='Houston').geo_filter(poly__contains='POINT(-95.362293 29.756539)')

Or combine both the bounding box routines (less accurate, fast) with the GEOS routines (most accurate, slower) to get a query that is both fast and accurate (this is not 'fast' in the current implementation, since geographic indices are not automatically created):

>>> qs = District.objects.geo_filter(poly__bbcontains='POINT(-95.362293 29.756539)').geo_filter(poly__contains='POINT(-95.362293 29.756539)')


Installation of the GeoDjango module will also require the installation of existing open source geographic libraries and a spatial database (currently only PostGIS). This section will describe the installation process for these libraries. Initially, these instructions will pertain only to a Linux platform (particularly Debian or Ubuntu). Mac & Windows support will be considered later; however, these instructions will most likely work through the Mac shell. Don't hold your breath for Windows support.


  • GeoDjango exists in the gis branch from SVN:
    $ svn co django_gis
    $ ln -s django_gis /path/to/site-packages/django


  • Latest GEOS version is 3.0.0RC4
    • Also requires SWIG >= 1.3.28. (Ubuntu Dapper comes with 1.3.27.)
    • If there's trouble locating your python, include PYTHON=/path/to/your/python.
      $ ./configure --enable-python
      $ make
      # make install


  • Latest PROJ.4 version is 4.5.0
    $ ./configure
    $ make
    # make install 
  • Should install datum shift files (for funky local coordinate systems) -- but I need to personally figure out how to do that first.


  • Latest PostGIS version is 1.2.1
  • First build & install PostGIS. We are currently using v8.1 of PostgreSQL.
    $ ./configure --with-geos --with-proj
    $ make
    # make install
  • Next, create a role and database for your application, and allow it to access PostGIS functionality:
    # su - postgres
    $ psql
    postgres=# CREATE ROLE <user> LOGIN;
    postgres=# \q
    $ createdb -O <user> <db_name>
    $ createlang plpgsql <db_name>
    $ psql -d <db_name> -f /usr/local/share/lwpostgis.sql
    $ psql -d <db_name> -f /usr/local/share/spatial_ref_sys.sql
    $ psql <db_name>
    <db_name>=# GRANT SELECT, UPDATE, INSERT, DELETE ON geometry_columns TO <user>;
    <db_name>=# GRANT SELECT ON spatial_ref_sys TO <user>;

  • Finally, update your to reflect the name and user for the spatially enabled database. So far, we only plan to support the psycopg2 backend, thus: DATABASE_ENGINE='postgresql_psycopg2'.


  • Optional, but highly useful for coordinate transformations and reading/writing both vector (e.g. SHP) and raster (e.g. TIFF) geographic data.
  • Latest GDAL version is 1.4.0. Configure with GEOS and Python support, then make and install:
    $ ./configure --with-geos --with-python
    $ make
    # make install
  • Note: This is done without the 'next generation' SWIG Python bindings. I've had trouble getting them to work, and the rumor is this only works on Windows. The compilation flag to enable these is --with-ngpython, but our packages currently only use the old bindings.

Model API


The following geometry-enabled fields are available:

  • PointField
  • LineStringField
  • PolygonField
  • MultiPointField
  • MultiLineStringField
  • MultiPolygonField
  • GeometryCollectionField

Field Keywords

  • Field keywords are used during model creation, for example:
    from django.contrib.gis.db import models
    class Zip(models.Model, models.GeoMixin):
      code = models.IntegerField()
      poly = models.PolygonField(srid=-1, index=True)
      object = models.GeoManager()
  • srid
    • Sets the SRID of geometry to the value. Defaults to 4326 (WGS84)
  • index
    • If set to True, will create a GiST index for the given geometry. Update the index with the PostgreSQL command VACUUM ANALYZE (may take a while to execute depending on how large your geographic-enabled tables are).

Creating and Saving Models with Geometry Fields

Here is an example of how to create a geometry object (assuming the Zip model example above):

>>> from zipcode.models import Zip
>>> z = Zip(code=77096, poly='POLYGON(( 10 10, 10 20, 20 20, 20 15, 10 10))')

Geometries are represented as strings in either of the formats WKT (Well Known Text) or HEXEWKB (PostGIS specific, essentially a WKB geometry in hexadecimal). For example:

Database API

Note: The following database lookup types can only be used with geo_filter(). All geographic queries are done with geo_filter() and geo_exclude(), thus separating the normal database API lookups from geographic-specific field queries. However chains containing both filter and geo_filter may still be used. Thus, geographic queries take the following form (assuming the Zip model used in the Model API section):

>>> qs = Zip.objects.geo_filter(<Zip geo field A>__<geo lookup type>=<geo string B>)
>>> qs = Zip.objects.geo_exclude(...)

PostGIS Operator Field Lookup Types

  • See generally, "Operators", PostGIS Documentation at Ch. 6.2.2
  • Note: This API is subject to some change -- we're open to suggestions.
  • overlaps_left
    • Returns true if A's bounding box overlaps or is to the left of B's bounding box.
    • PostGIS equivalent "&<"
  • overlaps_right
    • Returns true if A's bounding box overlaps or is to the right of B's bounding box.
    • PostGIS equivalent "&>"
  • left
    • Returns true if A's bounding box is strictly to the left of B's bounding box.
    • PostGIS equivalent "<<"
  • right
    • Returns true if A's bounding box is strictly to the right of B's bounding box.
    • PostGIS equivalent ">>"
  • overlaps_below
    • Returns true if A's bounding box overlaps or is below B's bounding box.
    • PostGIS equivalent "&<|"
  • overlaps_above
    • Returns true if A's bounding box overlaps or is above B's bounding box.
    • PostGIS equivalent "|&>"
  • strictly_below
    • Returns true if A's bounding box is strictly below B's bounding box.
    • PostGIS equivalent "<<|"
  • strictly_above
    • Returns true if A's bounding box is strictly above B's bounding box.
    • PostGIS equivalent "|>>"
  • same_as
    • The "same as" operator. It tests actual geometric equality of two features. So if A and B are the same feature, vertex-by-vertex, the operator returns true.
    • PostGIS equivalent "~="
  • contained
    • Returns true if A's bounding box is completely contained by B's bounding box.
    • PostGIS equivalent "@"
  • bbcontains
    • Returns true if A's bounding box completely contains B's bounding box.
    • PostGIS equivalent "~"
  • bboverlaps
    • Returns true if A's bounding box overlaps B's bounding box.
    • PostGIS equivalent "&&"

PostGIS GEOS Function Field Lookup Types

  • See generally "Geometry Relationship Functions", PostGIS Documentation at Ch. 6.1.2.
  • This documentation will be updated completely with the content from the aforementioned PostGIS docs.
  • distance
    • Return the cartesian distance between two geometries in projected units.
    • PostGIS equivalent Distance(geometry, geometry)
  • equals
    • Requires GEOS
    • Returns 1 (TRUE) if the given Geometries are "spatially equal".
    • Use this for a 'better' answer than '='. equals('LINESTRING(0 0, 10 10)','LINESTRING(0 0, 5 5, 10 10)') is true.
    • PostGIS equivalent Equals(geometry, geometry), OGC SPEC s2.1.1.2
  • disjoint
    • Requires GEOS
    • Returns 1 (TRUE) if the Geometries are "spatially disjoint".
    • PostGIS equivalent Disjoint(geometry, geometry)
  • intersects
    • PostGIS equivalent Intersects(geometry, geometry)
  • touches
    • PostGIS equivalent Touches(geometry, geometry)
  • crosses
    • PostGIS equivalent Crosses(geometry, geometry)
  • overlaps
    • PostGIS equivalent Overlaps(geometry, geometry)
  • contains
    • PostGIS equivalent Contains(geometry, geometry)
  • intersects
    • PostGIS equivalent Intersects(geometry, geometry)
  • relate
    • PostGIS equivelent Relate(geometry, geometry)

Extra Instance Methods

A model with geometry fields will get the following methods:


For every geometry field, the model object will have a get_FOO_wkt method, where FOO is the name of the geometry field. For example (using the School model from above):

>>> skool = School.objects.get(name='PSAS')
>>> print skool.get_point_wkt()
POINT(-95.460822 29.745463)


For every geometry field, the model object will have a get_FOO_centroid method, where FOO is the name of the geometry field. This routine will return the centroid of the geometry. For example (using the District model from above):

>>> dist = District.objects.get(name='Houston ISD')
>>> print dist.get_poly_centroid()
POINT(-95.231713 29.723235)


For every geometry field, the model object will have a get_FOO_area method, where {{{FOO}} is the name of the geometry field. This routine will return the area of the geometry.

>>> dist = District.objects.get(name='Houston ISD')
>>> print dist.get_poly_area()

Note: Units need to be figured out here.

Back to Top