2

after quite some time searching through SO, I did not find an example that solved my question:

How to label geom_bar(position = 'fill') using geom_text()?

mydf = data.frame(
  'sample' = sample(x = LETTERS[1:3], size = 111,replace = T),
  'category' = sample(x = 1:11, size = 111,replace = T)
)


mydf %>% 
  group_by(sample) %>%
  ggplot( aes(fill=sample, x=category)) + 
  geom_bar( position="fill" )
 # Does not work: +  geom_text(position='fill', aes(label = ..y..))

Intended outcome: label the % in the middle of the bars, as in the example (except that it shows the count, unlike needed in my case): from https://ggplot2.tidyverse.org/reference/geom_text.html

I guess the real question is what is "y" and "label" if I use geom_bar(position = 'fill'),

I am confused by ggplot blending calculations (proportions per category) with visualization.

Thanks!

bud.dugong
  • 689
  • 7
  • 16
  • 3
    There are already a lot of posts like this. Put `[ggplot2] label bar percent` in the SO search bar. Are you trying to do something that isn't covered in those? I'd also recommend doing the count and proportion calculations yourself before plotting, so you have control over them instead of relying on `geom_bar` to do it for you – camille Mar 14 '19 at 14:51

1 Answers1

6

Create a new data-frame ggp to store data for plotting

ggp <- mydf %>% count(sample,category)

Plot geom_bar using the data in data-frame - x=category, y=count(n), fill=sample

ggplot(ggp,aes(category,n,fill=sample)) +
geom_bar(stat="identity",position = "fill") + 
geom_text(aes(label=n),position = position_fill(vjust = 0.5))
bbiasi
  • 1,549
  • 2
  • 15
  • 31
Sithira Prabash
  • 76
  • 1
  • 4