When pulling census data as part of our predict_race call, the counties are split into groups by state. However, duplicates are left over. This is carried through in get_census_geo, causing the link for the initial county level pull to contain duplicated counties and the census data API to fail.
Added a line to call unique on each states' counties before passing it to the initial census data pull.
When pulling census data as part of our predict_race call, the counties are split into groups by state. However, duplicates are left over. This is carried through in get_census_geo, causing the link for the initial county level pull to contain duplicated counties and the census data API to fail.
Added a line to call unique on each states' counties before passing it to the initial census data pull.
This should fix #72