Proportion of 'Yes' Votes on Proposition 62 vs. log of County Population (11/08/2016)

prop62_pop_yes_prop.R

library(ggplot2)

votes <- read.csv('csv-ballot-measures.csv', header=TRUE, stringsAsFactors=FALSE)
votes <- votes[,-1:-2]
d.votes <- votes[which(votes$BALLOT_MEASURE_ID==62),]

drops <- c('COUNTY_ID', 'BALLOT_MEASURE_ID', 'BALLOT_MEASURE_NAME', 'BALLOT_MEASURE_TITLE')
d.votes <- d.votes[, !names(votes) %in% drops]

d.votes$YES_COUNT <- as.numeric(gsub(',', '', d.votes$YES_COUNT))
d.votes$NO_COUNT <- as.numeric(gsub(',', '', d.votes$NO_COUNT))
d.votes$POP <- d.votes$YES_COUNT + d.votes$NO_COUNT
d.votes$YES_PROP <- d.votes$YES_COUNT / d.votes$POP
d.votes$LOG_POP <- log10(d.votes$POP)

regr <- lm(YES_PROP ~ LOG_POP, data=d.votes)
regr.eqn <- paste0('y = ',
              toString(round(coefficients(regr)[1],2)),
              ' + ',
              toString(round(coefficients(regr)[2],2)),
              '*log(x)')
r2 <- paste0('R^2 = ', round(summary(regr)$adj.r.squared,2))

ggplot(data=d.votes, aes(x=LOG_POP, y=YES_PROP)) +
  geom_point() +
  geom_smooth(method=lm, se=FALSE, colour='red') +
  ggtitle('Prop 62 \'Yes\' Proportion vs. log(County Population)') +
  xlab('log(County Population)') +
  ylab('Proportion of \'Yes\' Votes') +
  geom_text(x=6.5, y=0.58, label=regr.eqn, parse=FALSE, colour='blue') +
  geom_text(x=6.5, y=0.55, label=r2, parse=FALSE, colour='blue')