How can I add individual data points to my bar plot using ggbarplot? (add jitter does not work, neither does dotplot)

Question

I am having a hard time adding individual data points on my barplots using ggbarplot. I seem to have an issue particularly on multi-grouped data (using fill or color). I've tried a couple different ways to add dots, but they've either resulted in an error (adding jitter) or the dots are not properly aligned to the correct bar (adding dotplot). Any help would be greatly appreciated.

Example data:

library(ggpubr)


#dataframe
ER <- read.table(header=TRUE, text="
  Sex  Group ER
  M    V     1046
  M    V     1290
  M    Z     1202
  M    Z     1056
  F    V     8000
  F    V     7859
  F    Z     4000
  F    Z     3409
")

I can get it to work if I use "add = c(point)" but when I try to use "add = c(jitter)" it results in an error ("position_jitterdodge()' requires at least one aesthetic to dodge by. I've seen similar questions asked here, but were usually resolved by adding the grouping factor to "Fill" instead of "color." I tried both and neither worked.

e <- ggbarplot(ER, x = "Sex", y = "ER", title = "Arcuate Nucleus",
                 ylab= "ER Labeling", xlab = "Group", color = "Group",
                 add = c("mean_se", "jitter"), add.params = list(color = "black"), palette = "jco",
                 position = position_dodge(.8)) 
print(e) 

Trying to use "add = c("dotplot)" results in the dots being mixed between my second group and shown in between the two side-by-side groups.

e2 <- ggbarplot(ER, x = "Sex", y = "ER", title = "Arcuate Nucleus",
                 ylab= "ER Labeling", xlab = "Group", color = "Group",
                 add = c("mean_se", "dotplot"), add.params = list(color = "black"), palette = "jco",
                 position = position_dodge(.8)) 
print(e2) 

This seems to work... but I would really like to use different colors, shapes, and add jitter. Example below.

e3 <- ggbarplot(ER, x = "Sex", y = "ER", title = "Arcuate Nucleus",
                 ylab= "ER Labeling", xlab = "Group", color = "Group",
                 add = c("mean_se", "point"), add.params = list(color = "black"), palette = "jco",
                 position = position_dodge(.8)) + theme(axis.line = element_line(size = 0.9), axis.ticks = element_line(size = 0.9)) 

print(e3)

Answer 1

The issue is that with add.params(color = "black") you are overwriting the grouping (color = "Group") needed for dodging the points. Hence, dropping add.params(color = "black") will make your code work. If you want black points and error bars you can do that via e.g. scale_color_manual by setting the colors for all groups to "black":

library(ggpubr)
#> Loading required package: ggplot2

ER <- read.table(header=TRUE, text="
  Sex  Group ER
  M    V     1046
  M    V     1290
  M    Z     1202
  M    Z     1056
  F    V     8000
  F    V     7859
  F    Z     4000
  F    Z     3409
")

ggbarplot(ER, x = "Sex", y = "ER", title = "Arcuate Nucleus",
               ylab= "ER Labeling", xlab = "Group", color = "Group", fill = "Group",
               add = c("mean_se", "jitter"), palette = "jco",
               position = position_dodge(.8)) +
  scale_color_manual(values = c(V = "black", Z = "black"))

EDIT

The issue is that packages like ggpubr offer a lot out of the box. The "price" one pays for that is that out-of-the-box solutions have their limits. If you want have more control about the appearence of your plot you have to do it (at least partially) using ggplot2, e.g. to color the points you can the jittered points using geom_jitter. Here I change the shape of the points so that I can add a black outline. Unfortunately it is not possibleto change the fill color as well as jitterdodge requires the fill aes:

ggbarplot(ER, x = "Sex", y = "ER", title = "Arcuate Nucleus",
          ylab= "ER Labeling", xlab = "Group", color = "Group", fill = "Group",
          add = c("mean_se"), palette = "jco",
          position = position_dodge(.8)) +
  scale_color_manual(values = c(V = "black", Z = "black")) +
  geom_jitter(aes(Sex, ER, fill = Group), shape = 21, color = "black",  position = position_jitterdodge(jitter.height = .1, jitter.width = .2))

Another option would be to use e.g.

geom_jitter(aes(Sex, ER, fill = Group), color = "grey",  position = position_jitterdodge(jitter.height = .1, jitter.width = .2))

which gives you "filled" e.g. "grey" points but with no outline.