0

I have produced a stacked bar plot using ggplot2 with the code below:

library(ggplot2)
g <- ggplot(data=df.m, aes(x=Type_Checklist,fill=Status)) +
 geom_bar(stat="bin", position="stack", size=.3) +             
 scale_fill_hue(name="Status") +    
 xlab("Checklists") + ylab("Quantity") +
 ggtitle("Status of Assessment Checklists") + 
 theme_bw() +
 theme(axis.text.x  = element_text(angle=90, vjust=0.5, size=10))+
 stat_bin(geom = "text",aes(label = paste(round((..count..)/sum(..count..)*100), "%")),vjust = 1)
 print(g)

The code manages to show the percentage as labels and maintained the actual quantities on the y axis. However, I wanted the percentages to be calculated per bar (shown on the x axis), not for the entire set of checks (bars in the plot).

I managed to do exactly that with the following code:

#count how many checklists per status`
   qty.checklist.by.status.this.week = df.m %>% count(Type_Checklist,Status)
   # Add percentage
   qty.checklist.by.status.this.week$percentage <- (qty.checklist.by.status.this.week$n/ nrNAs * 100)
   #add column position to calculate where to place the percentage in the plot
   qty.checklist.by.status.this.week <- qty.checklist.by.status.this.week %>% group_by(Type_Checklist) %>% mutate(pos = (cumsum(n) - 0.5 * n))


g <- ggplot(data=qty.checklist.by.status.this.week, aes(x=Type_Checklist,y=n,fill=Status)) +
 geom_bar(stat="identity",position="stack", size=.3) +             
 scale_fill_hue(name="Status") +
 xlab("Checklists") + ylab("Quantity") +
 ggtitle("Status of Assessment Checklists") +
 theme_bw() +
 theme(axis.text.x  = element_text(angle=90, vjust=0.5, size=10))+
 geom_text(aes(y = pos,label = paste(round(percentage), "%")),size=5)
 print(g)

But that solution seems quite manual, since I need to calculate the position of each label, different from the first plot that positions the labels automatically using stat_bin.

Existing answers have been very useful, such as Showing percentage in bars in ggplot and Show % instead of counts in charts of categorical variables and How to draw stacked bars in ggplot2 that show percentages based on group? and R stacked percentage bar plot with percentage of binary factor and labels (with ggplot) but they don't address exactly this situation.

To reiterate, how could I use the first solution, but calculate the percentages per bar, not for the entire set of bars in the plot?

Could anyone give me some help please? Thanks!

Community
  • 1
  • 1
Kenia Sousa
  • 31
  • 3
  • 2
    Generally speaking, you can't. The "right" solution, and the intended design philosophy of ggplot, is that you manipulate your data into the right format first, outside of ggplot, and then plot it. – joran Jul 08 '15 at 14:27
  • That's good to know about the philosophy. Just what was strange for me is that this second solution is even calculating the position where to show the percentage in the bar. But I didn't find any other solution so far. Thank you @joran ! – Kenia Sousa Oct 21 '15 at 09:25

0 Answers0