InterAction data

Summary: this document describes the per-country datasets on HDX derived from the InterAction's NGO Aid Map, via the HXL Proxy.

Source data overview

InterAction is an alliance of US-based iNGOs that publishes 3W-style data on some of their members' activities, used to drive their NGO Aid Map site. The data driving their map is available as a single, global dataset here:

https://www.ngoaidmap.org/downloads?doc=csv&name=12_projects&status=active

The data is also available as a set of per-country CSV (or KML) downloads, using a set of proprietary country codes.

https://www.ngoaidmap.org/downloads?doc=csv&geolocation=gn_3686110&level=1&name=colombia_projects&status=active

As it comes from the NGO Aid Map, the data includes four columns of personally-identifiable information (PII):

  1. project_contact_person
  2. project_contact_position
  3. project_contact_email
  4. project_contact_phone_number

HDX datasets

HDX has one dataset for each country covered by the NGO Aid Map, for example, InterAction member activities in Colombia. Each dataset includes two resources, one version of the data in CSV format, and one in JSON; both are generated by the HXL Proxy.

The Proxy reads the original CSV datasets, adds HXL hashtags to them, and then removes the four columns containing PII. See a copy of the data recipe for details.

The resources are all dynamic, generated whenever the user wants to download them (with a one-hour cache). At any given time, the resources for a dataset may or may not contain any activities, depending on the live data coming from InterAction.

Generating the datasets

The Python script used to generate or update the InterAction datasets is available in HDX's GitHub repository:

 https://github.com/OCHA-DAP/interaction-hdx 

Before running the script, you will need to copy the config.py.TEMPLATE file to config.py and fill in the appropriate values.

Note that script depends on a crosswalk between the internal NGO Aid Map country codes and HDX's ISO3 country codes, hosted in a publicly-readable Google Sheet:

https://docs.google.com/spreadsheets/d/1TfDOvNysztJCMLc0-EjnC2MpnYlpiVO0eio6XkT8V6I/edit?usp=sharing