Search code examples
rggplot2facet-wrap

How to order frequency counts of categorical variables by group facet_wrap in ggplot2?

I am trying to create bar graphs of categorical data ordered by frequency within groups. For example, using the diamonds dataset, I would like bar graphs ordered by counts/ frequency of "clarity" values grouped by "cut."

data(diamonds)
library(tidyverse)

diamonds$cut <- as.character(diamonds$cut)
diamonds$clarity <- as.character(diamonds$clarity)

diamonds %>% 
  group_by(cut) %>% 
  count(clarity, order = FALSE) %>% 
  mutate(clarity = as.factor(reorder(clarity, n))) %>% 
  ggplot(aes(n, clarity, fill = cut)) +
  geom_col(show.legend = FALSE) +
  facet_wrap(~ cut, ncol = 2, scales = "free") +
  labs(x = "count", y = NULL) +
  theme_bw()

This produces the following graph: enter image description here

It is very close to what I want, but the only graph actually ordered by clarity frequency is the first, "Fair." How can I make the all the bar graphs ordered by clarity frequency?

Thanks in advance.

Solution

  • One option would be to use reorder_within and scale_y_reorderd from the tidytext package:

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

diamonds$cut <- as.character(diamonds$cut)
diamonds$clarity <- as.character(diamonds$clarity)

diamonds %>%
  group_by(cut) %>%
  count(clarity, order = FALSE) %>%
  mutate(clarity = tidytext::reorder_within(clarity, n, cut)) %>%
  ggplot(aes(n, clarity, fill = cut)) +
  geom_col(show.legend = FALSE) +
  tidytext::scale_y_reordered() +
  facet_wrap(~cut, ncol = 2, scales = "free") +
  labs(x = "count", y = NULL) +
  theme_bw()