I have a data.frame in R 4.0.2 with a continuous variable in one column and two possible values of a categorical variable (variable 'type': known or novel) in another, which I use to color them differently (using a palette from ggsci 2.9 package). I represent an histogram (stat_bin) with ggplot2 3.3.2 and I want to use the facet_zoom function of ggforce 0.3.2 to zoom only the data belonging to one of the 'types' (using the option zoom.data, as it is done in the volcano example on http://cran.univ-paris1.fr/web/packages/ggforce/vignettes/Visual_Guide.html#contextual-zoom), however I get this error:
Error: Aesthetics must be either length 1 or the same as the data (2000): x
Reproducible example:
library(ggplot2)
library(ggsci)
library(ggforce)
testdata <- as.data.frame(sort(rnorm(1000)))
testdata$type <- "known"
testdata[501:1000,2] <- "novel"
# Working code
ggplot(testdata) +
stat_bin(aes(x=testdata[,1], fill = type), binwidth = 1, color="white") +
scale_fill_npg() + theme_light() +
facet_zoom(xlim = c(0, 4), ylim = c(0, 300), horizontal = TRUE, zoom.size = 0.3)
# Desired code
ggplot(testdata) +
stat_bin(aes(x=testdata[,1], fill = type), data = cbind(testdata, zoom = FALSE), binwidth = 1, color="white") +
stat_bin(aes(x=testdata[testdata$type == "novel",1]), data = cbind(testdata, zoom = TRUE), binwidth = 0.5) +
scale_fill_npg() + theme_light() +
facet_zoom(xlim = c(0, 4), ylim = c(0, 300), horizontal = TRUE, zoom.size = 0.3, zoom.data = zoom)
Thanks!
The issue is that you pass the whole dataset as data in the second
stat_bin
. Simply pass the subsetted df instead of trying to subset inaes()
:BTW: I also renamed the first variable in your data as x.