• cablab - Public CAB-LAB API
• cablab.cube - Data Cube Generation and Access (protected, public parts expr)
• cablab.cube_cli - Command-line interface (protected)
• cablab.util - Common utility functions (protected)

Find the CAB-LAB documentation here.

Adhere to PEP-8!

Only place TODOs in source code when you have an according issue on GitHub. Mention the issue number in the TODO text. When fixing a TODO, mention the issue in the commit message.

Test code in the test and test/providers directories should only use libraries that are anyway used by the production code in src. If you want to check out new libraries for appropriateness please do so in the test/sandbox directory.

• cablab.image_providers: key = class derived from cablab.ImageProvider

• cablab.image_providers:
  • 'burnt_area = cablab.providers.burnt_area.BurntAreaProvider'
• console_scripts:
  • 'cablab_cli = cablab.cli:main', see %PYTHON_HOME%/Scripts/cablab_cli (*.exe on Windows) after installation

Development mode installation:

> python setup.py develop

or real installation:

> python setup.py install

Create a file cablab-config.py in your project root directory or your current working directory and add the following entry:

cube_sources_root = <your local cube source directory>

To generate a default data cube with a 0.25 degree resolution and variables 'BurntArea', 'C_Emmisions', Ozone', 'Precip' call the cube-gen tool:

> cube-gen testcube burnt_area:dir=BurntAreaDir c_emissions:dir=EmissionsDir ozone:Ozone-CCI/Total_Columns/L3/MERGED precip:dir=CPC_precip

It's usage is:

> cube-gen <cube-dir> [<provider-key>:dir=<source-path> ...]

If you use Windows, get the Python wheels from Christoph Gohlke's website at http://www.lfd.uci.edu/~gohlke/pythonlibs/. Install them using:

> pip install <wheel-file>

• netCDF4 >= 1.2
• numpy >= 1.9
• gridtools