Geographic IDs in case data

[smit1678]

This was originally captured in an email and discussed on the phone, but moving to here to track. Case data, and any data, needs to have geographic IDs associated with it. Naming inconsistencies cause conflicts when joining with geographic data.

To keep this consistent, we should use WHO admin codes to match with WHO geo data.

cc @pendergast @jpolonsky @drewbo @dereklieu

[smit1678]

New data is nearly ready. Combined shapefiles of admin1 admin2 here: https://dl.dropboxusercontent.com/u/29368453/COUNTRIES_ADM1_2.zip. Waiting on Admin0.

[drewbo]

Let's treat these new shapefiles as the standard for both boundaries and centroids; the file we were sent earlier has odd definitions for ADM1/2 in Liberia compared to every other data set we've seen, including this one