The datahub is a platform for people to share data and to cooperate on data refinement and management. The site allows users to share simple resources (e.g. a file or an API endpoint address) and to bundle those into datasets: groupings for multiple represenatations of a single logic dataset, lists of resources on a shared topic or a list of files required for a common purpose. Each resource can be part of multiple datasets, so the same reference can be used in multiple contexts.

To install datahub, you need to load its dependencies. For that, you may want to create a virtualenv before you perform these steps:

pip install -r pip-requirements.txt
pip install -e .

To configure datahub, create a copy of datahub/default_settings.py with appropriate configuration settings. When starting datahub, set the environment variable DATAHUB_SETTINGS to the path of your local config file.

You also need to make sure that an elastic search daemon is running at the location specified in your configuration file. Elastic search is required even if you just want to run the tests.

If you also want to run asynchronous tasks, you will need to create a file called celeryconfig.py, normally by symlinking either to your local configuration or to datahub/default_settings.py.