cloud-harness: Build custom GBDX tasks for beginners

PyPI version

This package allows a user to build custom GBDX tasks that can be executed locally and remotely on the GBDX platform. All from a python file and commandline tools.

To run remote tasks, this package uses the gbdx-auth package. Therefore the users credentials must be configured in the ~/.gbdx-config, see here for more details.

For general information on the GBDX platform and API, see the GBDX University.

Remote Runtime

When running tasks remotely, the runtime environment is pre-configured with a number of common science, data, and image processing libraries. The following is a list of what is available:

  • GDAL 1.10.1
  • Cython 0.24
  • Boto 2.40.0
  • OpenCV 2.4.8
  • Numpy 1.8.2
  • Scipy 0.13.3
  • Matplotlib 1.5.1
  • Scikit-learn 0.17.1
  • PyProj 1.9.5.1
  • Scikit-image 0.12.3
  • Protobuf 2.6.1
  • Imutils 0.3.6
  • Yapps 2.2.0
  • Profilehooks 1.8.1
  • Fiona 1.6.4
  • Shapely 1.5.15
  • Descartes 1.0.2
  • Pillow 2.9.0
  • Psycopg2 2.6.1
  • Jinja2 2.8
  • Awscli 1.10.28
  • Gbdxtools 0.2.6

Some of the versions above are not the latest. If you require a newer version please create an issue on Github here: gbdx-cloud-harness