Closed kuriwaki closed 1 week ago
note: remember to incorporate the unrelated magnitude fix https://github.com/kuriwaki/cvr_harvard-mit_scripts/pull/347#issue-2498321226 when re-running the entire pipeline
I looked through and approve everything with this PR, merge in when you are ready @kuriwaki .
Dropped the ballot style variable and re-ran.
This PR:
Revises the redaction policy from countywide to precinct-specific.
The (untracked) file has one row per each
precinct_cvr
which contains one unanimous office. It then crosswalks that revealed precinct toprecinct_cvr
, which is a set of more than oneprecinct_cvr
s so that the group is unanimous.The edit uses this crosswalk and runs
coalesce(precinct_group, precinct)
. Same forprecinct_medsl
To consider: if a revealed precinct has no appropriate match, the
precinct_cvr_group
is calledUNIQUE REVEALED STYLE
. Replace this value with NA to retain the original precinct_cvr.