Reputation: 4386
I have some data, and I want to use some variables from stat_count()
to label a bar plot.
This is what I want to do:
library(ggplot2)
library(scales)
percent_and_count <- function(pct, cnt){
paste0(percent(pct), ' (', cnt, ')')
}
ggplot(aes(x=Type)) +
stat_count(aes(y=(..prop))) +
geom_text(aes(y=(..prop..), label=percent_and_count(..prop.., ..count))),
stat='count')
However, I get this error, since it can't find the function in what I assume is either some base packages or the data frame:
Error in eval(expr, envir, enclos) : could not find function "percent_and_count"
I get this error if I do percent(..prop..)
as well, although it is fine with scales::percent(..prop..)
. I did not load my function from a package.
If all else fails, I can do
geom_text(aes(y=(..prop..), label=utils::getAnywhere('percent_and_count')$objs[[1]]((..prop..),(..count..))))
But this seems needlessly roundabout for what should be a stupidly simple task.
Upvotes: 0
Views: 484
Reputation: 50668
You can use bquote
and aes_
:
# Sample data
set.seed(2017);
df <- data.frame(
Type = sample(6, 100, replace = T)
);
library(ggplot2);
library(scales);
# Your custom function
percent_and_count <- function(pct, cnt){
paste0(percent(pct), ' (', cnt, ')')
}
ggplot(df, aes(x = Type)) +
stat_count(aes(y = ..prop..)) +
geom_text(
stat = "count",
aes_(
y = ~(..prop..),
label = bquote(.(percent_and_count)((..prop..), (..count..)))))
Explanation: bquote(.(percent_and_count)(...))
ensures that percent_and_count
is found (as terms .(...)
are evaluated in the parent environment). We then use aes_
to ensure that quoted expressions (either with ~
or bquote
) are properly evaluated.
Still not pretty, but probably more straighforward than using utils::getAnywhere
.
Upvotes: 1