I have three factors and have produced the following plot

   mtcars%>% ggplot( aes(as.factor(carb),fill=as.factor(am)))+
    geom_bar(aes( y=..count../tapply(..count.., ..fill.. ,sum)[..fill..]), 
             position="fill")+
    facet_wrap(~as.factor(vs))+
    theme(legend.position = "bottom")

I can get what I want side by side but I'd like to produce a version where both plots are in one. Plot I’m aiming for in picture below (bottom plot).

Originals were made using python but cannot find any ggplot2 code to reproduce.

I have three factors and have produced the following plot

   mtcars%>% ggplot( aes(as.factor(carb),fill=as.factor(am)))+
    geom_bar(aes( y=..count../tapply(..count.., ..fill.. ,sum)[..fill..]), 
             position="fill")+
    facet_wrap(~as.factor(vs))+
    theme(legend.position = "bottom")

I can get what I want side by side but I'd like to produce a version where both plots are in one. Plot I’m aiming for in picture below (bottom plot).

Originals were made using python but cannot find any ggplot2 code to reproduce.

• r
• ggplot2
• plot

1 Answer 1

You can create this kind of chart easily by using the negative values for the category to be shown on the left. To make things a bit easier I would suggest to first compute the proportions manually outside of ggplot. Finally, to add the category labels you can use annotate where I use coord_cartesian(clip="off") to draw outside of the panel borders and added some more margin at the top to make room for the labels:

library(ggplot2)
library(dplyr, warn = FALSE)

mtcars |>
  count(carb, am, vs) |>
  mutate(prop = n / sum(n), .by = c(carb, vs)) |>
  mutate(prop = if_else(vs == 0, -1, 1) * prop) |>
  ggplot(aes(
    y = as.factor(carb), x = prop,
    fill = as.factor(am)
  )) +
  geom_col(color = "white") +
  scale_x_continuous(labels = ~ scales::percent(abs(.x))) +
  theme(legend.position = "bottom") +
  annotate(
    x = c(-Inf, Inf),
    y = Inf,
    label = c("vs = 0", "vs = 1"),
    geom = "label",
    hjust = c(0, 1),
    vjust = 0,
    fill = NA,
    label.size = 0,
    fontface = "bold"
  ) +
  coord_cartesian(clip = "off") +
  theme(
    plot.margin = margin(22, 5.5, 5.5, 5.5)
  )