2016 primary election

README.md

###### chart 1. 2016 Hawaii primary election results.

pairing national election data and maps is a hot mess.

Recently, I have come across a few choropleth maps using 2016 primary election data presenting relative proportions of voters by county for a single candidate. One example that stands out is a map of Bernie supporters in Alaska. The heavily colored map feels misleading when considering ~600 democrats participated. More people could fit in a Walmart on Black Friday.

Sometimes, a less sexy visualization tells an interesting story. My modest stacked bar graph shows the proportion of Republican and Democratic voters (portrays Hawaii’s huge imbalance towards the Democratic party) and the relative proportion of votes each candidate garnered.

Choropleth maps can be beautiful and effective, especially when making state-by-state comparisons. Other times, less fancy charts can be just as effective. It’s all about the story.

cover.png

src.R

# hawaiikernel.com
# primary election

# INTRODUCTION -----------------------------------------------------------------
# primary election chart

# DEPENDENCIES -----------------------------------------------------------------
# http://www.politico.com/2016-election/results/map/president/hawaii

dir.rtools <- "%PATH%;C:/RBuildTools/3.3/gcc-4.6.3/bin;c:/RBuildTools/3.3/bin"
Sys.setenv(PATH=dir.rtools)

setwd("D:/scratch")

pacman::p_load(dplyr, lubridate, rvest, stringr, tidyr, readr, purrr, jsonlite)

# DATA -------------------------------------------------------------------------
primary_HI_R <- fromJSON("http://data.cnn.com/ELECTION/2016primary/HI/county/S.json", 
                         flatten=TRUE)
primary_HI_D <- fromJSON("http://data.cnn.com/ELECTION/2016primary/HI/county/E.json", 
                         flatten=TRUE)
primary_HI_R$counties

pacman::p_load(lazyeval)

scrapeJSON <- function(df, candidate) {
  which_column <- "lname"
  filter_criteria <- interp(~which_column==candidate, which_column=as.name("lname"))
  
  y <- mutate(map_df(df$counties$race.candidates, function(x) {
    x %>% filter_(filter_criteria)}),
    FIPS=df$counties$countycode,
    COUNTY=df$counties$name)
  return(y)
}

cruz    <- scrapeJSON(primary_HI_R, "Cruz")
kasich  <- scrapeJSON(primary_HI_R, "Kasich")
trump   <- scrapeJSON(primary_HI_R, "Trump")
sanders <- scrapeJSON(primary_HI_D, "Sanders")
clinton <- scrapeJSON(primary_HI_D, "Clinton")

# PREPROCESS -------------------------------------------------------------------

republican <- rbind_list(cruz, kasich, trump) %>%
  group_by(FIPS) %>%
  mutate(N=sum(votes),
         P=votes / sum(votes)) %>%
  replace(is.na(.), 0)

democrat <- rbind_list(sanders, clinton) %>%
  group_by(FIPS) %>%
  mutate(N=sum(votes),
         P=votes / sum(votes)) %>%
  replace(is.na(.), 0)

primary_df <- rbind_list(republican, democrat)

primary_df <- primary_df %>%
  select(lname, party, COUNTY, P, votes) %>%
  mutate(party=ifelse(party=="D", "Democratic Party",
                        ifelse(party=="R", "Republican Party", NA)))

answers_order <- c("Clinton", "Sanders", "Trump", "Cruz", "Kasich")

get_middle_position <- function(middle_pos, levels, start_values, end_values) {
  if(middle_pos %% 1 != 0) {
    
    if(ceiling(middle_pos) %in% levels) {
      value <- (end_values[levels==ceiling(middle_pos)] - start_values[levels==ceiling(middle_pos)]) * .5 + start_values[levels==ceiling(middle_pos)]
    } else {
      value <- start_values[levels > middle_pos][1]
    }
  } else {

    if(middle_pos %in% levels){
      value <- end_values[levels==middle_pos]
    } else {
      value <- start_values[levels > middle_pos][1]
    }
  }
  value
}

middle_pos <- 2

primary_df2 <- primary_df %>%
  ungroup %>%
  mutate(COUNTY=as.factor(COUNTY),
         lname=factor(lname, levels=answers_order)) %>%
  arrange_(.dots=c("COUNTY", "lname"))

primary_df_RP <- primary_df2 %>%
  group_by_(.dots=c("COUNTY")) %>%
  filter(!is.na(lname)) %>%
  mutate(response_relative=votes / sum(votes),
         y_start=lag(cumsum(response_relative), default=0),
         y_end=y_start + response_relative,
         middle=get_middle_position(middle_pos, as.numeric(lname), y_start, y_end)) %>%
  mutate(y_start_center=y_start - middle,
         y_end_center=y_end - middle,
         y_max_centered=max(y_end_center)) %>%
  ungroup

primary_df_wide <- primary_df2 %>%
  group_by_(.dots=c("COUNTY")) %>%
  summarise(missing=length(lname[is.na(lname)]),
            total=n(),
            completed=total - missing) %>%
  left_join(primary_df_RP %>%
              group_by_(.dots=c("COUNTY")) %>%
              summarise(neg=abs(min(y_start_center)),
                        pos=max(y_end_center)) %>%
              ungroup, by=c("COUNTY"))

primary_df_long <-
  primary_df_wide %>%
  gather(type, value, completed, missing)

final <- list(
  percentages=primary_df_wide,
  main=primary_df_RP %>% filter(!is.na(lname)),
  counts=primary_df_long)

# CHART ------------------------------------------------------------------------
pacman::p_load(ggplot2)
windowsFonts(Consolas=windowsFont("Consolas"))

png("primary.png",height=500,width=960)

ggplot(final$main) +
  geom_rect(aes(xmin=as.numeric(COUNTY)-.4, xmax=as.numeric(COUNTY)+.4,
                ymin=y_start_center, ymax=y_end_center, fill=lname)) +
  geom_hline(yintercept=0, lty=2) +
  geom_text(aes(x=as.numeric(COUNTY), y=-1.06, label=paste0(round(neg*100), "%")),
            data=final$percentages,size=6, fontface="bold", colour="#535353") +
  geom_text(aes(x=as.numeric(COUNTY), y=1.06,  label=paste0(round(pos*100), "%")), 
            data=final$percentages, size=6, fontface="bold", colour="#535353") +
  coord_flip() +
  scale_y_continuous("", breaks=seq(-1,1,.5), limits=c(-1.1,1.1), 
                     labels=c("100%","50%","0","50%","100%")) +
  xlab("County") + ylab("Votes") +
  theme(legend.position="bottom") +
  scale_x_continuous(breaks=1:length(levels(final$main$COUNTY)), 
                     labels=levels(final$main$COUNTY)) +
  scale_fill_manual("",
                    labels=c("Clinton", "Sanders", "Trump", "Cruz", "Kasich"),
                    values=c("#9ecae1", "#3182bd", "#de2d26",
                               "#fc9272", "#fee0d2")) +
  ggtitle(expression(
    atop("State of Hawaii 2016 Primary Election Results",
         atop("Primary election results by County"), ""))) +
  theme_bw() +
  theme(panel.background=element_rect(fill="#F0F0F0"),
        plot.background =element_rect(fill="#F0F0F0"),
        panel.border    =element_rect(colour="#F0F0F0"),
        panel.grid.major=element_line(colour="#D0D0D0",size=.75),
        panel.grid.minor=element_blank(),
        axis.ticks=element_blank(),
        legend.position="right",
        legend.key=element_blank(),
        legend.background=element_blank(),
        strip.text.x=element_text(face="bold", family="Consolas", 
                                  size=8, colour="#3C3C3C"),
        plot.title=element_text(face="bold", family="Consolas", vjust=-1, 
                                colour="#3C3C3C",size=22),
        axis.text.x=element_text(size=16, family="Consolas", 
                                 colour="#535353",face="bold", angle=90),
        axis.text.y=element_text(size=12, family="Consolas", 
                                 colour="#535353",face="bold"),
        axis.title.y=element_text(size=16, family="Consolas",
                                  colour="#535353",face="bold",vjust=0),
        axis.title.x=element_text(size=16, family="Consolas",
                                  colour="#535353",face="bold",vjust=0)) +
  geom_hline(yintercept=21.20894,size=1.2,colour="#535353")

dev.off()

thumbnail.png