1

I was trying to create a stacked bar plot using ggplot2 with labels showing the cumulative sum of a variable. But the bars are stacked in the wrong order so the label does not match the bars.

The data set is cabbage_exp from the gcookbook package.

The process is shown below.

I first used arrange to order the data by Cultivar first, then by Date. Then I used ddply to create a new column label_y to reflect the cumulative sum of Weight, grouped by Cultivar.

ce <- arrange(cabbage_exp, Cultivar, Date)

ce <- ddply(ce, "Cultivar", transform, label_y = cumsum(Weight))

After the above two steps, the data looks like this.

  Cultivar Date Weight        sd  n         se label_y
1      c39  d16   3.18 0.9566144 10 0.30250803    3.18
2      c39  d20   2.80 0.2788867 10 0.08819171    5.98
3      c39  d21   2.74 0.9834181 10 0.31098410    8.72
4      c52  d16   2.26 0.4452215 10 0.14079141    2.26
5      c52  d20   3.11 0.7908505 10 0.25008887    5.37
6      c52  d21   1.47 0.2110819 10 0.06674995    6.84

Then I created the barplot using the following code.

ggplot(data=ce, aes(x=Cultivar, y=Weight, fill=Date)) + 
geom_bar(stat="identity") + geom_text(aes(y=label_y, label=label_y), 
vjust=1.5, colour="white")

The diagram looks like this. The order of the bars for d16, d20 and d21 is not correct.

enter image description here

I then tried to change the order of dates and regenerated the graph using the following code. But the graph did not change.

ggplot(data=ce, aes(x=Cultivar, y=Weight, fill=Date, order=desc(Date))) + 
geom_bar(stat="identity") + geom_text(aes(y=label_y, label=label_y), 
vjust=1.5, colour="white")

I am not sure where I went wrong. Any help will be greatly appreciated.

Thanks very much.

neilfws
  • 32,751
  • 5
  • 50
  • 63
Di Wu
  • 13
  • 2
  • I managed to find a solution by modifying the order of **Date** by using `ce <- arrange(cabbage_exp, Cultivar, desc(Date))`. This way, the bar plots at least correspond to the cumulative sum labels. However, by doing this, among *d16*, *d20* and *d21*, *d16* has the largest cumulative sum, which is not very logical. – Di Wu Apr 21 '17 at 02:02
  • Make `Date` a factor with the levels in the desired order. – alistaire Apr 21 '17 at 02:26
  • If you want the stack order to be from bottom to top instead of top to bottom, see `position = position_stack(reverse = TRUE)` in `geom_bar`. – aosmith Apr 21 '17 at 02:28
  • Thanks @aosmith. It worked perfectly! Thanks so much. – Di Wu Apr 21 '17 at 06:53

1 Answers1

0

The reason your labels don't correspond to the category is that you are plotting the label based on a y intercept, not based on category correspondence. So the highest y-intercept value always comes at the top.

So you just need to call the label as part of the ggplot(aes... call to say "for each category, put this corresponding text on it". 

#try this

ggplot(data = ce, aes(x = Cultivar, y = Weight, fill = Date, label = label_y)) + 
  geom_bar(stat = "identity") + 
  geom_text(size = 3, position = position_stack(vjust = 0.5)) #play with position to control where the text is placed.

what you want

(also, would encourage you to practice good code indentation to improve readability)

griffmer
  • 357
  • 2
  • 9
  • Thanks for the suggestion @griffmer. It kinda works but because the labels represent cumulative sum, it would be better if from bottom stack to the top stack, the cumulative sum increases. The suggestion provided by aosmith worked perfectly :) – Di Wu Apr 21 '17 at 06:52