Page MenuHomePhabricator

Add new region definitions to canonical data repository
Closed, DuplicatePublic

Description

Courtesy of @gpaumier, we now have an official set of Wikimedia regions. We should add these to the canonical-data repository.

Steps:

Event Timeline

We will need to meet with GDI to get a consensus on their team's mapping of country codes and various labels and regional groupings implemented in T310712

kzimmerman added subscribers: EChetty, CMacholan.

For now we are using the GDI mapping for ad hoc analysis. I think we need to have a meeting with @EChetty and @CMacholan / GDI to clarify our use cases and work on how to bring the regional view into Superset dashboards. However, several other issues are higher priority at the moment so I am lowering the priority of this task.

We are waiting for the GDI tables to be productionized before moving forward with this. Need to check in on T310712

GDI's work will be tracked in T324681
I have asked @JAnstee_WMF to provide details of the db and table name so we're aware.

nshahquinn-wmf moved this task from Waiting on others to Next 2 weeks on the Movement-Insights board.
nshahquinn-wmf added a subscriber: kzimmerman.

An update from @JAnstee_WMF on July 6:

I worked with Community Resources & Guillaume P. to resolve final differences between grant committee and foundation regional alignments and pushed final changes to production gdi.country_meta_data table. The only difference between foundation_region and grant_committee_region columns now is that the Middle East & Africa committee covers both the Middle East & Northern Africa region and SubSaharan Africa region; all country-to-region associations are otherwise now aligned across schemas.

So it's now safe to add these definitions to canonical_data.countries.

@JAnstee_WMF also wrote that "Kate mentioned we still need to implement some version controls." I think that would be satisfied by naming the field something like "wmf_region_2023". That way, if we ever update our region definitions, we avoid having the earlier definition look like the canonical one. I think I'll invite some opinions on Slack to see if people would find this useful.