I need to create several boxplots describing my data. I have an annoying X axis which is continuous and I want to keep it true to scale, but at the same time I need an axis break as the gap is very large between the two largest numbers on the X axis.
So far (very simplified) I have something like this:
library(ggplot2)
library(ggbreak)
yrs<-c("2", "8", "17", "21", "24","64")
df <- data.frame(treatm = factor(rep(c("A", "B"), each = 18)),
seeds = (c(sample.int(1000, 36, replace = TRUE))),
years= as.numeric(rep(yrs), each = 6))
ggplot(df, aes(x = years, y = seeds, fill = treatm, group= interaction(years,treatm))) +
geom_boxplot() +
scale_x_continuous(breaks = c(2,8,17,21,24,64),
labels = paste0(yrs))+
scale_x_break(c(26, 62)) +
theme_classic()
which creates ugly graph
ggbreak is the only way I have managed to create a break in the X axis, but of course the result is far from optimal. Most importantly, I need to remove the strange X axis which appears on top! Optimally, I would like to include some sort of axis break symbol on the axis, like a little zigzag or two dashes or something. Currently I'm finishing up the graphs in Photoshop but that is an enormous waste of time so please help!!
To hide the top x axis you can add the following line :
theme(axis.text.x.top = element_blank(),
axis.ticks.x.top = element_blank(),
axis.line.x.top = element_blank())
To include an axis break symbol, the only "solution" I see is to manually place a tag (// or ... for example) in your graph.
labs(tag = "//") +
theme(plot.tag.position = c(0.9, 0.1))
The two parameters of the tag position are values between 0 and 1. They vary depending on the size of your output image so you have adjust it yourself.
My solution applied to your example gives the following code :
library(ggplot2)
library(ggbreak)
yrs<-c("2", "8", "17", "21", "24","64")
df <- data.frame(treatm = factor(rep(c("A", "B"), each = 18)),
seeds = (c(sample.int(1000, 36, replace = TRUE))),
years= as.numeric(rep(yrs), each = 6))
ggplot(df, aes(x = years, y = seeds, fill = treatm, group= interaction(years,treatm))) +
geom_boxplot() +
scale_x_continuous(breaks = c(2,8,17,21,24,64),
labels = paste0(yrs))+
scale_x_break(c(26, 62)) +
theme_classic()+
theme(axis.text.x.top = element_blank(),
axis.ticks.x.top = element_blank(),
axis.line.x.top = element_blank())+
labs(tag = "//") +
theme(plot.tag.position = c(0.767, 0.15)) # <-- manually adjusted parameters