I am frustrated from ggplot2 overwriting arguments that have been set before. In fact I encouter the described problem with different ggplot2 functions, but in order to illustrate what I mean I write only about scale_y_continuous(). So let’s say I provide a different breaks argument for scale_y_continuous() as here:

df <- data.frame(sex= factor(sample(c("male", "female"), 100, replace= TRUE)))
library(ggplot2)
my_plot <- ggplot(df, aes(sex)) +
  geom_bar() +
  scale_y_continuous(breaks= 1:nrow(df))
my_plot

This returns the expected plot:

Let’s assume we want to change the y axis to start at zero. But if I add another argument in scale_y_continuous(), then the former breaks argument is changed back to default. See here:

my_plot2 <- my_plot +
  scale_y_continuous(expand= expansion(mult= c(0, 0.1)))
my_plot2

How can I hinder ggplot2 from overwriting the breaks argument that has been set before?

Why I ask: We have a function that produces different plots for all our data situations. This function does a bunch of things, one can provide different data or plots and it returns plots in the layout as we need them. The function itself is too big to share here. Anyway, it is frustrating that ggplot2 sets defined arguments back to default.