Position dodge for aes fill and position stack for aes color

Asked Apr 17 '19 at 09:41

Active Apr 17 '19 at 09:41

Viewed 63 times

I would like to plot this data frame with a barplot :

df <- data.frame(x = rep(c("x1", "x2"), each = 4L), 
                 f = rep(c("f1", "f2"), each = 2L, times = 2L),
                 c = rep(c("c1", "c2"), 4L),
                 y = c(rbind(seq(0.1, 0.4, by = 0.1), 1 - seq(0.1, 0.4, by = 0.1))))

I would like to dodge the bars for distinct fill values but also keep them stacked for the same fill values, when it's just the color value which is different.

library(ggplot2)
ggplot(df, aes(x = x, 
               y = y, 
               fill = f, 
               color = c)) + 
  geom_col(position = "dodge")

So in this case all bars would be stacked to 1 with distinct "color" in it.

Do you have any idea on how I can achieve this ? I would like to avoid faceting.

Thank You !

edited Jun 20 '20 at 09:12

Community

asked Apr 17 '19 at 09:41

Julien Navarre

7,653
3
42
69

maybe adding `group = f` to `aes`? – amatsuo_net Apr 17 '19 at 09:48
Like so `ggplot(df, aes(interaction(x, f), y, value, fill = c)) + geom_col()` ? – markus Apr 17 '19 at 09:50
Relevant: [ggplot2 - bar plot with both stack and dodge](https://stackoverflow.com/questions/12715635/ggplot2-bar-plot-with-both-stack-and-dodge) – markus Apr 17 '19 at 09:52
@markus I saw this thread but I ideally don't want to use `interaction` to mofidy my x values – Julien Navarre Apr 17 '19 at 10:17
@amatsuo_net how would you use the aes(group = f), I can't make it work on my side – Julien Navarre Apr 17 '19 at 11:22
1

Can you elaborate on *why* you'd want to avoid both `facet_*` & `interaction()`? – Z.Lin Apr 20 '19 at 04:44

Position dodge for aes fill and position stack for aes color

0 Answers0