Beatbox is a simple Python wrapper for GDAL, Shapely, GeoRasters, and GeoPandas. It is designed to be a portable interface for working with large spatial datasets on cloud platforms like AWS and Google App Engine / Earth Engine. You install it's standard dependencies with anaconda or pip and can produce raster and vector datasets that can be integrated into common spatial processing frameworks and distributed computing workflows. PLJV uses beatbox downstream for cloud computing tasks. You will see it used as a standard dependency on a number of our other projects. Beatbox supports Python 2 and 3 out-of-the-box, but you will need to use Python 2 for Earth Engine functionality.
This is still a young project. If you are not a collaborator and just happened
onto this repository while looking for ways to crunch geospatial data with
Python, this is probably not your tool. There are more mature upstream
projects available. For vector data, consider
GeoPandas.
For raster data, consider GeoRasters. For
distributed computing workflows,
consider GeoTrellis.
For an overview of installing the Earth Engine Python API (RTM).
We maintain beatbox and make it publicly available (GPLv3) so that our collaborators can check our math. It is still very much under active development and not production ready. For the curious earth systems modeller experimenting with spatial data on Earth Engine or AWS that would like to get their hands dirty, dive in.
From a direct download:
python setup.py install
From conda / pip:
conda install pyCrypto GDAL numpy pandas fiona shapely geopandas scikit-learn
pip install oauth2client
pip install google-api-python-client
pip install earthengine-api
pip install --upgrade git+git://github.com/PLJV/Beatbox.gitfrom beatbox import Do, Vector, Raster, fuzzy_convex_hulls
water_raster = Raster("/path/to/water_raster.tif")
spatial_points = Vector("/path/to/spatialpoints.shp")
convex_hulls = fuzzy_convex_hulls(spatial_points, width=1033)
result = Do({
'what': extract,
'with': [ convex_hulls, water_raster ]
})