Changelog • ccesMRPprep

Add CD-level analyses

Separate out the synth_functions to synthjoint package (#9)

Overwrite cd_info so it uses two party vote (it makes a big difference in Utah in 2016). Add 2016 and 2020 lines too.
Allow ACS up to 2019 or error out on unavailable years
Update get_cces_dataverse to 2020 CES with validated vote
Add option to make White Hispanics as “Hispanics” and make it a default; same for Black Hispanics (806beb2)

Fix to wrong values in get_acs_cvap (#7)

Introduce synth_prod(), synth_mlogit(), synth_smoothfix()
Add New York data as a third sample (acs_NY, cc18_NY, elec_NY)
Add option for states in get_acs()

Rename get_cces_dv to get_cces_dataverse()
Use CRAN version of dataverse (#3)
Change formula notation in build_counts and cces_join_slim to (y ~ x) instead of yes | trials(n_response)

Update cces_dv_ids link to new Dataverse versions (1ac26f993aa88e343ad686175d885c88f393e3ca)
Make Native American a proper category (1f6d8d0b08d5b2e5afb196150b8f79cc54209957)
Add a CPS vignette (c64528ee96f136b592c128e7424eab2bea0e9fc8)

Change default to ACS-1yr instead of ACS-5yr, since the Census ACS-5 yr should not be used if we want to prioritize recent values (ad30b4c6e181833801ab0604da69849e493e7c57). I guess ACS-5yrs do not up weight more recent years.
Use Crunch terminology of names and values for the labels (ce08f874d211bcc89c9ad9a379b039144d50dbba)

allow for different names and additional variables in count, get_cces_question, cces_join_slim (a59ad9b105255927365cd59162fc1a4ba4ec1188, a59ad9b105255927365cd59162fc1a4ba4ec1188)

Change default so that age gets replaced to bins in ccc_std_demographics
treat Don’t knows as NA in yesno_to_binary

Sample CCES 2018
Derived Variables Vignette
More sample questions

Added a NEWS.md file to track changes to the package.