Saving ggplot objects for use on different datasets

Question

We all know we can do

p <- ggplot(data)
p + aes(x = funny_hat) + geom_histogram()

But I've never seen documentation of the opposite case. Say I want to store a bunch of ggplot objects, e.g.:

geom_boxplot() + stat_compare_means(method = "anova", size = 14, vjust = 1) +
scale_y_continuous(breaks = seq(-2,2.5, 0.5)) + theme(text = element_text(size = 48))

... so that I can plug all of that in for different datasets e.g.

ggplot(data1) + [all my arguments]
ggplot(data2) + [all my arguments]

This doesn't work if I try to use the same method as I would in the first example, but is there a way to do it? I grow tired of copy-pasting...

Answer 1

The trick is use a list of the elements to make a template, and then apply the template to different data frames:

library(ggplot2)
library(ggpubr) 

plot_template <- list(
  geom_boxplot(),
  stat_compare_means(method = "anova", size = 14, vjust = 1),
  scale_y_continuous(breaks = seq(-2, 2.5, 0.5)),
  theme(text = element_text(size = 48))
)

# test
ggplot(mtcars, aes(x = factor(gear), y = mpg)) + plot_template

ggplot(iris, aes(x = Species, y = Petal.Length)) + plot_template

Answer 2

You can do this with the (underused/underappreciated) %+% operator, which substitutes a new data object and rebuilds the object.

Building on @TarJae's example,

gg0 <- ggplot(iris, aes(x = Species, y = Petal.Length)) + plot_template
gg0 %+% mtcars +  aes(x = factor(gear), y = mpg)

(if you were doing this with a bunch of data sets that all had the same variables you wouldn't need to add the aes(...) to override the previous mappings ...)