CODEWITHSUNDEEP
rmatrixggplot2histogrambar-chart
einisieuza
einisieuza

Reputation: 1

Stacked Barplot in R using ggplot and a Matrix of a likert scale

this is my first question here. I've searched for an answer before, but I wasn't able to find a satisfying one - probably because I am an absolute beginner with R.

I have the following data:

      pbw1 pbw2 pbw3 pbw4 pbw5 pbw6
[1,]   2   3   2   1   0   2    3
[2,]   2   1   1   3   4   5    6

PBW is a variable answered on a likert scale. This dataset has 1014 rows, so this is just an example.

What I need is a figure that looks like that:

Example

For my dataset this means: 6 items (I'd like to replace pbw with a sentence in the end) on the x-axis and therefore 6 bars. These should show how many persons of the n=1014 (in percent) have answered the item. Like, 80% have answered with 0, 10% answered with 1, 5% answered with 5%, etc. This should obviously be cumulative.

All I read is ggplot, melt, and so on. Yet I am just unable to get R to do what I want and need.

This didn't help:

datm = melt(cbind(pbwmatrix, ind=rownames(pbwmatrix)), id.vars=c('ind'))

Also barplot(pbwmatrix) did not; also several silly combinations of random commands did not (yes, I am desperate).

Help in any way would be much appreciated!

Upvotes: 0

Views: 1195

Answers (1)

lukeA
lukeA

Reputation: 54287

Here's a starter using ggplot:

df <- as.data.frame(replicate(6, sample(1:6,100,T)))
library(tidyverse)
df %>% 
  gather %>% 
  group_by(key, value) %>% 
  tally %>% 
  mutate(n = n/sum(n)*100) %>% 
  ggplot(aes(x=key, y=n, fill=as.factor(value))) + 
  geom_col() + 
  geom_text(aes(label=n), position=position_stack(.5)) + 
  coord_flip()

Upvotes: 2

Related Questions