polettif / proporz Goto Github PK

Content:

• finish #7
• Cleanup dataset documentation, add references
• Combine documentations (pivot_to_*, divisor_methods, quorum_functions)
• update method list and links in README

First release:

• usethis::use_news_md()
• update news.md
• usethis::use_cran_comments()
• Update (aspirational) install instructions in README
• Proofread Title: and Description:
• Check that all exported functions have @return and @examples
• Check that Authors@R: includes a copyright holder (role 'cph')
• add author role to Authors@R
• Check licensing of included files
• Review https://github.com/DavisVaughan/extrachecks

Prepare for release:

• git pull
• usethis::use_github_links()
• urlchecker::url_check()
• devtools::check(remote = TRUE, manual = TRUE)
• devtools::check_win_devel()
• git push

Submit to CRAN:

• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()

I feel like there is something off with the party divisors:

The seats allocated to each party differ more than they should from their share of votes. Please see attached the data I used.
votes_FIN_2019.xlsx
district_seats_FIN_2019.xlsx

Using a 0.5 % national treshold, e.g. party "RKP" should land 9 seats, but using pukelsheim only gives them 7.
"VAS" should be entitled to 17 due to their share, but they get 18 etc.

The specs I use:

seats_df = pukelsheim(votes_df,
                      district_seats_df,
                      quorum_districts = 0,
                      quorum_total = 0.005)

Here is how precise results I got by finding the divisors by hand:
result_Pukelsheim_by_hand_FIN_2019.xlsx

Am I missing something? Thanks for your patience!

Add a "winner takes at least one" option (Majorzbedingung) to biproportional and pukelsheim.

The following input data leads to an error in biproportional.R#252:

library(proporz)
df_votes = structure(
    list(list_id = c("SDP", "PS", "KOK", "PS", "KOK", "SDP", "KOK", "SDP", "PS"),
         entity_id = c("HEL", "UUS", "HAEM", "HEL", "UUS", "HAEM", "HEL", "UUS", "HAEM"),
         list_votes = c(4000, 9000, 17000, 23000, 29500, 36000, 42500, 49000, 17999)),
    row.names = c(NA, -9L), class = "data.frame")
df_district_seats = structure(
    list(entity_id = c("HEL", "UUS", "HAEM"),
         election_mandates = c(5, 10, 15)),
    row.names = c(NA, -3L), class = "data.frame")

pukelsheim(df_votes, df_district_seats)
#> Error in bisect(fun, divisor_range[1], divisor_range[2], tol): 
#> (f(x1) <= 0 & f(x2) >= 0) | (f(x1) >= 0 & f(x2) <= 0) is not TRUE

^{Created on 2020-07-22 by the reprex package (v0.3.0)}

While working with your library, I found an issue calculating seat allocation for Zurich's cantonal elections. The canton of Zurich sets the quorum at either 0.05 in at least 1 district or 0.03 overall. The pukelsheim function, however, does not allow for that. It would be great if you could add this option to the function.

Thank you for the amazing work!

Prepare for release:

• git pull
• Check current CRAN check results
• Polish NEWS
• urlchecker::url_check()
• devtools::check(remote = TRUE, manual = TRUE)
• devtools::check_win_devel()
• Update cran-comments.md

Submit to CRAN:

• usethis::use_version('minor')
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()

Add vignettes for the readme examples and especially the analysis done in finland-comparison.md