R - ggplot2 - plotting one factor by two factors

Question

I'm following up on this link: ggplot2 bar plot with two categorical variables

My problem is very similar – I need to do what the OP needed but I need to split the count bars by another factor.

Let me illustrate:

>     var response_cat has values 0,1 (categories of a response) 
>     
>     var group indicates belonging to a group (0,1) 
>     
>     var item indicates the name of an item (20 strings with codes like LY1, GN6,...)

And I need to draw a graph as follows: x axis has two bars per item (for each group) y axis has relative frequencies of ones

any ideas?

thank you!

then you may need `facet_wrap` Try `df %>% group_by_at(names(.)) %>% summarise(n = n()) %>% ggplot(., aes(x = Fruit, y = n, fill = Bug)) + geom_bar(stat = "identity") + facet_wrap(~ group)` — akrun, May 05 '18 at 16:52
Just to clarify – I'm not the OP! But thank you very much, I'm going to try it — Petr Palíšek, May 05 '18 at 17:10
If you are trying from that link try with `group <- rep(LETTERS[1:2], each = 4); df <- data.frame(Fruit,Bug, group)` — akrun, May 05 '18 at 17:11
the result was a bit strange, I suppose the NAs might be at fault?: https://imgur.com/a/zygIz6u — Petr Palíšek, May 05 '18 at 17:16
You are right. But "as.factor" doesn't seem to change it, which is weird — Petr Palíšek, May 05 '18 at 17:32
As you didn't provide an example, it is not clear to me. I am suggesting from the dataset in that link — akrun, May 05 '18 at 17:33
that part of the dataframe only has 0, 1 and NA, if that helps — Petr Palíšek, May 05 '18 at 17:44
May be try to change that `NA` to "Others" and check if that works by converting to `factor` — akrun, May 05 '18 at 17:45
I solved it by re-importing the data set and forcing the "factor" and your solution works! Thank you — Petr Palíšek, May 05 '18 at 17:46

score 3 · Accepted Answer · edited Jun 20 '20 at 09:12

We could use facet_wrap

library(tidyverse)
df %>%
   group_by_at(names(.)) %>%
   summarise(n = n()) %>%
   ggplot(., aes(x = Fruit, y = n, fill = Bug)) + 
   geom_bar(stat = "identity") +
   facet_wrap(~ group)

data

Fruit <- c(rep("Apple",3),rep("Orange",5))
Bug <- c("worm","spider","spider","worm","worm","worm","worm","spider")
group <- rep(LETTERS[1:2], each = 4)
df <- data.frame(Fruit,Bug, group)

R - ggplot2 - plotting one factor by two factors

1 Answers1

data