R/hare_niemeyer.R

In coalitions: Bayesian "Now-Cast" Estimation of Event Probabilities in Multi-Party Democracies

#' Seat Distribution by Hare/Niemeyer
#'
#' Calculates number of seats for the respective parties that have received more
#' than \code{hurdle} percent of votes (according to the method of Hare/Niemeyer)
#'
#' @param votes Number of votes per party.
#' @param parties Names of parties (must be same length as votes).
#' @param n_seats Number of seats in parliament. Defaults to 183 (seats in
#' Austrian parliament).
#' @seealso \code{\link{sls}}
#' @return A \code{data.frame} containing parties above the hurdle and the respective
#' seats/percentages after redistribution via Hare/Niemeyer
#' @examples
#' library(coalitions)
#' library(dplyr)
#' # get the latest survey for a sample of German federal election polls
#' surveys <- get_latest(surveys_sample) %>% tidyr::unnest("survey")
#' # calculate the seat distribution based on Hare/Niemeyer for a parliament with 300 seats
#' hare_niemeyer(surveys$votes, surveys$party, n_seats = 300)
#' @export
hare_niemeyer <- function(votes, parties, n_seats = 183) {

  quotas <- votes / sum(votes) * n_seats
  seats <- floor(quotas)
  rest <- quotas - seats
  seats_left <- n_seats - sum(seats)
  for (i in 1:seats_left) {
    m_i <- which.max(rest)
    if (length(m_i) > 1)
      m_i <- sample(m_i, 1)
    seats[m_i] <- seats[m_i] + 1
    rest[m_i] <- 0
  }
  seats

}