R/voter_profile.R

In electionsBR: R Functions to Download and Clean Brazilian Electoral Data

#' Download data on the voters' profile
#'
#' \code{voter_profile()} downloads and cleans data on the voters' profile aggregated by state, city and electoral zone.
#' The function returns a \code{data.frame} where each observation corresponds to a voter profile type.
#'
#' @param year Election year (\code{integer}). For this function, the following years are available: 1994, 1996, 1998,
#' 2000, 2002, 2004, 2006, 2008, 2010, 2012, 2014, 2016, 2018 and 2020.
#' 
#' @param ascii (\code{logical}). Should the text be transformed from Latin-1 to ASCII format?
#'
#' @param encoding Data original encoding (defaults to 'windows-1252'). This can be changed to avoid errors
#' when \code{ascii = TRUE}.
#' 
#' @param export (\code{logical}). Should the downloaded data be saved in .dta and .sav in the current directory?
#' 
#' @param temp (\code{logical}). If \code{TRUE}, keep the temporary compressed file for future use (recommended)
#'
#' @details If export is set to \code{TRUE}, the downloaded data is saved as .dta and .sav
#'  files in the current directory.
#'
#' @return \code{voter_profile()} returns a \code{data.frame} with the following variables:
#'
#' \itemize{
#'   \item PERIODO: Election year.
#'   \item UF: Units of the Federation's acronym in which occurred the election.
#'   \item MUNICIPIO: Municipality name.
#'   \item COD_MUNICIPIO_TSE: Municipal's Supreme Electoral Court code (number).
#'   \item NR_ZONA: Electoral zone's Supreme Electoral Court code (number).
#'   \item SEXO: Voters' sex.
#'   \item FAIXA_ETARIA: Voters' age group.
#'   \item GRAU_DE_ESCOLARIDADE: Voters' education degree.
#'   \item QTD_ELEITORES_NO_PERFIL: Absolute number of voters.
#' }
#'
#' @import utils
#' @importFrom magrittr "%>%"
#' @export
#' @examples
#' \dontrun{
#' df <- voter_profile(2002)
#' }

voter_profile <- function(year, ascii = FALSE, 
                          encoding = "windows-1252",
                          export = FALSE,
                          temp = TRUE){
  
  
  # Inputs
  if(!year %in% seq(1994, 2020, by = 2)) stop("Invalid 'year'. Please check the documentation and try again.")
  test_encoding(encoding)
  
  #if(year == 2020){
  #  urldir <- "http://agencia.tse.jus.br/estatistica/sead/odsele/perfil_eleitorado/perfil_eleitorado_ATUAL.zip"
  #} else{
  #   urldir <- sprintf("http://agencia.tse.jus.br/estatistica/sead/odsele/perfil_eleitorado/perfil_eleitorado_%s.zip", year) 
  #  }
  
  filenames  <- paste0("/perfil_eleitorado_", year, ".zip")
  dados <- paste0(file.path(tempdir()), filenames)
  url <- "https://cdn.tse.jus.br/estatistica/sead/odsele/perfil_eleitorado%s"
  
  # Downloads the data
  download_unzip(url, dados, filenames, year)
  
  # remover temp file
  if(temp == FALSE){
    unlink(dados)
  }
  
    
  setwd(as.character(year))
  
  archive <- Sys.glob("*")[grepl(".pdf", Sys.glob("*")) == FALSE] %>%
    file.info() %>%
    .[.$size > 200, ] %>%
    row.names()
  
  if(grepl(".csv", archive[1])){
    test_col_names <- TRUE
  }else{
    test_col_names <- FALSE
  }
  
  banco <- readr::read_delim(archive, col_names = test_col_names, delim = ";", locale = readr::locale(encoding = encoding), col_types = readr::cols(), progress = F) %>%
    dplyr::as_tibble()
  
  setwd("..")
  unlink(as.character(year), recursive = T)

  
  # Change variable names
  if(year < 2016){
    names(banco) <- c("PERIODO", "UF", "MUNICIPIO", "COD_MUNICIPIO_TSE", "NR_ZONA",
                      "SEXO", "FAIXA_ETARIA", "GRAU_DE_ESCOLARIDADE", "QTD_ELEITORES_NO_PERFIL")
  } 
  
  
  # Change to ascii
  if(ascii) banco <- to_ascii(banco, encoding)
  
  # Export
  if(export) export_data(banco)
  
  message("Done.\n")
  banco
}