In cotecora-team-2/quickcountmx: Quick-count estimation package for the 2021 Mexican election

knitr::opts_chunk$set(echo = TRUE)

reescalar <- TRUE
is_median_ylim <- FALSE

library(tidyverse)
get_ylim <- function(df,is_median) {
  if(is_median)
    max((df %>% group_by(candidato) %>% summarise(m = median(median,na.rm = TRUE)) %>% filter(!is.na(m)))$m)*1.05
  else
    max((df %>% group_by(candidato) %>% summarise(m = max(median,na.rm = TRUE)) %>% filter(!is.na(m)))$m)*1.05
}

path_buzon2 <- "/home/rstudio/workspace/cotecora_s14/buzon2/"
estimaciones_files <- list.files(paste0(path_buzon2,"estimaciones"), full.names = TRUE, recursive = TRUE, include.dirs = FALSE, pattern = "*csv")
razon_files <- list.files(paste0(path_buzon2,"razon"), full.names = TRUE, recursive = TRUE, include.dirs = FALSE, pattern = "*csv")
pctpropobs_files <- list.files(paste0(path_buzon2,"pctpropobs"), full.names = TRUE, recursive = TRUE, include.dirs = FALSE, pattern = "*csv")
multi_files <- estimaciones_files[grepl("ortizm|anzarut",estimaciones_files,fixed = FALSE)]
mendoza_files <- estimaciones_files[grepl("mendoza",estimaciones_files,fixed = TRUE)]
compulsado_files <- estimaciones_files[grepl("compulsado",estimaciones_files,fixed = TRUE)]

get_df <- function(edo_id){
    out <- tryCatch(
        {
          get_fdate <- function(x){
            x <- as.character(x)
            nx <- as.POSIXct(paste0(substr(x,1,2)," ",substr(x,3,4),":",substr(x,5,6)),format = "%d %H:%M")
           }
          candidatos <- readr::read_csv("../data-raw/estados_candidatos_partidos_2021.csv") %>%
            filter(ID_ESTADO == as.numeric(edo_id)) #%>%
    #    filter(!grepl("IC",CANDIDATO)) #quita candidatos independientes
          lista_candidatos <- candidatos$CANDIDATO %>% unique()
          lista_candidatos <- lista_candidatos[!grepl("OTROS", lista_candidatos, fixed = TRUE)]
          lista_candidatos <- append(lista_candidatos,"PART")

          file_edos <- substr(multi_files, nchar(multi_files) - 11,nchar(multi_files) - 10)
          multi_df <- multi_files[file_edos == edo_id] %>% map_df(., ~ read_csv(.) %>%
                                                              mutate_at(vars(EN), ~
                                                                          as.character(.)) %>%
                                                              mutate(method="multi"))
          file_edos <- substr(mendoza_files, nchar(mendoza_files) - 11,nchar(mendoza_files) - 10)
          mendoza_df <- mendoza_files[file_edos == edo_id] %>% map_df(., ~ read_csv(.)  %>%
                                                              mutate(method="mendoza"))
          file_edos <- substr(razon_files, nchar(razon_files) - 11,nchar(razon_files) - 10)
          razon_df <- razon_files[file_edos == edo_id] %>% map_df(., ~ read_csv(.)) %>%
                                                              mutate(PART=as.numeric(""))  %>%
                                                              mutate(method="razon")
          file_edos <- substr(pctpropobs_files, nchar(pctpropobs_files) - 5,nchar(pctpropobs_files) - 4)
          pctpropobs_df <- pctpropobs_files[file_edos == edo_id] %>% map_df(., ~ read_csv(., col_names = c("EN","R","pctpropobs"),col_types = "cdd")) 
          file_edos <- substr(compulsado_files, nchar(compulsado_files) - 11,nchar(compulsado_files) - 10)
          compulsado_df <- compulsado_files[file_edos == edo_id] %>% map_df(., ~ read_csv(.)  %>%
                                                              mutate(method="compulsado") %>%
                                                                select(-ESTRATOS,-EST_REC,-TOT_CAS,-CAS_REC,-PORCENTAJE))
          df <- do.call("rbind", list(multi_df,mendoza_df,razon_df,compulsado_df))
          df <- df %>%
            ungroup() %>%
            left_join(pctpropobs_df, by=c("EN","R")) %>%
            tidyr::pivot_longer(cols = all_of(lista_candidatos), names_to = "candidato",
                 values_to = "porcentaje") %>%
            rename(fecha = R) %>%
            mutate_at(vars(fecha), get_fdate) %>%
            select(-EQ) %>%
            unique() %>%
            group_by(EN,fecha,pctpropobs) %>%
            mutate(LMU = dplyr::case_when(
            LMU == 0 ~ "inf",
            LMU == 1 ~ "median",
            LMU == 2 ~ "sup")) %>%
            tidyr::pivot_wider(names_from = LMU, values_from = porcentaje) 

          df
        },
        error=function(cond) {
            message(paste("No hay datos para",edo_id))
            message(cond)
            return(NA)
        }
    )    
    return(out)
}

df6 <- get_df("06")
df8 <- get_df("08")
df16 <- get_df("16")
df18 <- get_df("18")
df32 <- get_df("32")

plotDF_edo <- function(df,edo_str) {
    out <- tryCatch(
        {
          limite <- ifelse(reescalar, get_ylim(df,is_median_ylim), 100)
          ann_text<-data.frame(fecha=df$fecha,
                               y=rep(limite ,nrow(df)),
                               candidato=df$candidato,
                               label=substr(as.character(signif(df$pctpropobs,digits=2)),1,3))
          ann_text <- ann_text %>% mutate(label=ifelse(as.numeric(substr(as.character(fecha),15,16)) %% 15, "",label)) %>% mutate(label=ifelse(label=="NA","",label))
          ggplot(df, aes(fecha,median, color=factor(candidato))) +
            geom_line(aes(group = candidato)) +
            geom_errorbar(aes(ymin = inf, ymax = sup), width = 0.5) +
            ylim(0,limite) +
            facet_wrap(~method, nrow=1) +
            xlab("Hora") +
            ylab("Porcentaje") +
            labs(subtitle = paste("Votacion para gobernador de",edo_str)) +
            theme(legend.position="bottom") +
            geom_text(data = ann_text,aes(x=fecha,y=y,label=label), size = 3, alpha = 0.1, inherit.aes = FALSE)
        },
        error=function(cond) {
            message(paste("No hay datos para",edo_str))
            message(cond)
            return(NA)
        }
    )    
    return(out)
}

plotDF_edo(df6,"Colima")
plotDF_edo(df8,"Chihuahua")
plotDF_edo(df16,"Michoacan")
plotDF_edo(df18,"Nayarit")
plotDF_edo(df32,"Zacatecas")

cotecora-team-2/quickcountmx documentation built on July 17, 2025, 5:14 a.m.