Admin unit indexing
clim2parquet provides an index file for administrative units that aims to follow the GADM format, allowing users to link prepared climate data with GADM data.
The package data admin_units_index is a dataset with the following columns taken from GADM:
-
"GID_0" (country ISO 3 character code);
-
"GID_1", "GID_2", "GID_3" (sub-national unit identifiers);
-
"NAME_0", "NAME_1", "NAME_2", "NAME_3" (country and sub-national unit names).
A further column "admin_unit_id" provides an integer identifier for each admin unit.
Differences from GADM data
The admin unit index file differs from GADM data, and the column "admin_unit_id" does not correspond to the "UID" column found in GADM data.
There are extra entries (rows) for each admin level < 3, that is entries where for "GID_X" every "GID_(Y < X)" is missing.
In GADM data, each entry is resolved down to the lowest available admin level (which may be as low as level 5 for some countries), with a unique identifier "UID" for each one.
Since we resolve data down to multiple levels of interest (up to level 3), we have had to construct a separate identifier that is applicable only to climate data generated using clim_to_parquet.
You will need the package admin unit index file in order to link prepared climate data with GADM data.
Preparing the index file
The index file is prepared from GADM spatial data. See and run the script data-raw/admin_unit_ids.py to generate data. The recommended way of doing this is using the uv package manager.