# the output of your binomial logistic model
# will return log odds.
# 
# Here's a function to convert that number to a proportion
# (which is probably how your data are summarized)


convert_logodds_to_prop <- function(log_odds){
  proportion <- 1/(1+exp(-log_odds))
  return(proportion)
}

# examples:
convert_logodds_to_prop(0)
# log odds of zero converts to 0.5 proportion

convert_logodds_to_prop(1)
# log odds of 1 equals proportion of 0.731

convert_logodds_to_prop(-1)
# log odds of -1 equals proportion of 0.269

convert_logodds_to_prop(13)
# log odds of 13 is extremely close to 1.0 !

# make a plot
df <- data.frame(logodds = seq(-4, 4, 0.025))
df$proportion = convert_logodds_to_prop(df$logodds)

px_logodds <- ggplot(df)+
  aes(x = logodds, y = proportion)+
  geom_line()+
  xlab("Log odds")+
  ylab("Proportion")
px_logodds