Simple raster file implementation for Django based on PostGis
Django-raster provides the simplest possible raster data integration for Django projects using a PostGIS database backend.
Note: This package requires a PostGIS >= 2.0
Install package with pip install django-raster
Add “raster” to your INSTALLED_APPS setting like this:
INSTALLED_APPS = ( ... 'raster', )
Run python manage.py migrate to create the raster models.
(Optional) Add RASTER_USE_CELERY = True to your project’s setting to enable asynchronous raster parsing.
Django-raster provides the simplest possible integration of raster data in Django. It is based on the python bindings provided by the GDAL package. Raster files can be uploaded and parsed through the admin interface. The raw raster data can be parsed asynchronously if Celery is integrated into the Django project (see below).
Once a raster file is uploaded, the parser will extract the data in the raster file and store the rasters in regular tiles of 100x100 pixels in a PostGIS raster table. Each tile will be one row in a PostGIS raster table.
For this, the package defines two models and one field:
- RasterLayer - storing the raw raster files and meta-data (for example rasterfile=raster.tif and srid=4326)
- RasterTile - storing the parsed raster in PostGis. The raster data is split into tiles of 100x100 pixels and each tile is stored as an instance of RasterTile. The raster data itself is stored in a RasterField within the RasterTile model.
- RasterField - an extension of the Django BinaryField class to store the raster data. The only difference to the BinaryField class is that the db_type is redefined as raster, such that the data is stored as raster in PostGIS.
Due to the simplicity of the implementation, no spatial querying can be done on the raster data through python. This package is not integrated with GeoDjango and has none of the features that GeoDjango spatial models have. Proper integration of rasters into Geodjango with spatial querying etc. is beyond the scope of this package. If required however, custom SQL can be used to make spatial queries on the raster data.
After setting the package up, you can upload raster files through the admin interface using the RasterLayer model. Specify a layer name, the raster data type (continuous, categorical, mask or rank ordered), the raster’s srid, the nodata value and the raster file to be uploaded.
Upon saving the raster file, django-raster automatically loads the raster data from the file into a raster column in the PostGIS database. These tiles will be stored in the RasterTile model, which should not be edited directly but only through adding and deleting entire RasterLayers instances.
The RasterLayer instances have a parse_log field, which stores information about the parsing process. For debugging, there might be some useful information in the parse log.
Asynchronous parsing with Celery
For large rasters files the parsing step might take a while, so the html request that stored the raster might time out. To avoid this, django-raster can easily integrated with Celery .
To use celery for the raster parsing step, which is triggered automatically after saving RasterLayer instances, add the following setting to your django settings file:
RASTER_USE_CELERY = True
If this setting is enabled, Celery pushes one task to the queue for each raster that is saved. The raster is then parsed upon execution of the task by a worker. The default of this setting is False.
The default tile size is 100x100 pixels. The tile size can be changed by providing an integer value in the RASTER_TILESIZE setting. The tiles are always saquares, so the tileize is set by one integer that specifies the number of pixels in each tile. For instance, setting:
RASTER_TILESIZE = 200
will import the raster in tiles of 200x200 pixels. Note that if you change this setting after uploading rasters, the tile size will not be updated on the existing rasters.
By default, the tiles on the edge of the raster file are padded such that all raster tiles for one rasterlayer are of the same size. If you dont want the raster tiles to be padded at the edges of the raster, you can disable padding through the following setting:
RASTER_PADDING = False
Release history Release notifications
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size django-raster-0.1.0.tar.gz (10.9 kB)||File type Source||Python version None||Upload date||Hashes View hashes|