GA and LA TIGER TABBLOCK20 files fail to load for years 2020 through 2023

[weyhing]

Both Georgia and Louisiana have malformed TABBLOCK20 shapefiles (for 2020 through 2023 download years). These two states will not load into tiger_data tables inherited from tiger.tabblock20.

The problem is that each shapefile is missing a UATYPE20 column, and also has an unexpected GEOIDFQ20 column. These are the only two TABBLOCK20 shapefiles with this problem.

Considering the UATYPE20 column is as-of-yet unused I've worked around the problem by adding a couple lines of code to the ga_load.cmd and la_load.cmd scripts:

%PSQL% -c "ALTER TABLE tiger_staging.la_tabblock20 ADD COLUMN IF NOT EXISTS uatype varchar(1) null;"
%PSQL% -c "ALTER TABLE tiger_staging.la_tabblock20 DROP COLUMN IF EXISTS geoidfq20;"

Specific TIGER files tested:

www2.census.gov\geo\tiger\TIGER2020\TABBLOCK20\tl_2020_13_tabblock20.zip
www2.census.gov\geo\tiger\TIGER2020\TABBLOCK20\tl_2020_22_tabblock20.zip

www2.census.gov\geo\tiger\TIGER2022\TABBLOCK20\tl_2022_13_tabblock20.zip
www2.census.gov\geo\tiger\TIGER2022\TABBLOCK20\tl_2022_22_tabblock20.zip

www2.census.gov\geo\tiger\TIGER2023\TABBLOCK20\tl_2023_13_tabblock20.zip
www2.census.gov\geo\tiger\TIGER2023\TABBLOCK20\tl_2023_22_tabblock20.zip

See attached output for specific dtypes using geopandas.

[weyhing]

Output from Jupyter notebook showing malformed TABBLOCK20 shapefiles in comparison with state of Michigan.

[robe]

Since it's only two states, sounds like an issue with tiger data distributed which they might have fixed already.

Also I never promise that a 2023 load script will work for prior versions, or maybe you are using the right year load script for those.

That said I don't think this issue is worth fixing unless it is an issue in 2024 data as well.

I'll make sure to test the 2024 for louisiana and Georgia before I close out #5810