Humdata

A humanitarian data service. The aspirational goal is to consolidate the various fragmented raw data sources, whether it be APIs, libraries, raw urls, or PDFs, behind a single REST API and expos via Swagger UI. This is the initial effort to lay the groundwork for this to be possible.

Preliminary TODOs (by 3/10)

• install all dependencies and track them in reqs.txt files (e.g. swagger)
• use hdx api to download, parse, and merge some data (first focus on lake chad basin)
• lay out structure
• lay out skeleton for rest api and swagger ui
• add landing page with links to api docs and repo
• construct endpoints for accessing the data
• a single data source end to end
• a set of lake chad basin data sources end to end (more granular than country level, or historic data)

Swagger UI

See the website with interactive API documentation here. Note: this is currently local, to see it run the following:

python api.py

Current set of endpoints (for Lake Chad Basin, 2016-2017):

• GET /funding/totals/:country
• GET /funding/categories/:country
• GET /needs/totals/:country
• GET /needs/regions/:country

Raw data sources

To pull data from HDX (the Humanitarian Data Exchange), run the following:

python3 run_hdx.py

This data script is configured to run every Monday at 2:30am (system time) for the latest data. See resources/constants.py and resources/data/raw:

• HDX Lake Chad Basin Key Figures January 2017
• HDX Lake Chad Basin FTS Appeal Data
• HDX Lake Chad Basin Crisis Displaced Persons
• Lake Chad Basin Humanitarian Needs Overview January 2017
• UNOCHA ORS ROWCA

Derived data sources

See resources/data/derived - this cleaned and formatted data is what the API is ultimately serving.