ggplot2 : Plot mean with geom_bar

Solution 1:

simply use stat = "summary" and fun.y = "mean"

ggplot(test2) + 
  geom_bar(aes(label, X2, fill = as.factor(groups)), 
           position = "dodge", stat = "summary", fun.y = "mean")

Solution 2:

ggplot2 likes 1 data point for 1 plot point. Create a new data frame with your summary statistics, then plot with stat="identity"

require(reshape2)
plot.data <- melt(tapply(test2$X2, test2$groups,mean), varnames="group", value.name="mean")

 ggplot(plot.data, aes(x=group,y=mean)) + geom_bar(position="dodge", stat="identity")

Solution 3:

Try using ggpubr. It creates ggplot2-like charts.

library(ggpubr)

ggbarplot(test2, x = "label", y = "X2",
          add = "mean", fill = "groups")

Alternatively, add a facet:

ggbarplot(test2, x = "label", y = "X2",
          add = "mean", fill = "groups",
          facet.by = "groups")