Different tick length on the same axis in ggplot2

Question

I've seen two great examples on SO where two sets of different ticks were added to the same plot in ggplot2, see Insert blanks into a vector for, e.g., minor tick labels in R and ggplot2 displaying unlabeled tick marks between labeled tick marks. However, what if I want two sets of ticks with different lengths? It's fairly easy to do this in base R (data and code modified from ref 2):

library("magrittr")
library("ggplot2")

set.seed(5)
df <- data.frame(x = rnorm(500, mean = 12.5, sd = 3))

breaks <- seq(2.5, 25, .5)

plot(hist(df$x,breaks = breaks), xaxt = "n", col = "gray66")
axis(1, tck = -.02, at = breaks[breaks %% 2.5 == 0], lwd = 2, lwd.ticks = 2)
axis(1, tck = -.01, lwd = 0, at = breaks[breaks %% 2.5 != 0], labels = NA, lwd.ticks = 1)

and I get (notice the two sets of ticks on X axis with different lengths):

base R plot

I don't see how this is done in ggplot2, the axis.ticks.length arguement in theme() only takes the first element of a vector for plotting when I tried passing a vector of the same length of the breaks.

Answer 1

As Peter mentionned there is a ggh4x package for that:

install.packages('ggh4x')
library(ggh4x)

set.seed(5)
df <- data.frame(x = rnorm(500, mean = 12.5, sd = 3))


    ggplot(df,aes(x=x) )+
      geom_histogram()+
      scale_x_continuous(
        minor_breaks = seq(0, 20, by = 1),
        breaks = seq(0, 20, by = 5), limits = c(0, 20),
        guide = "axis_minor" # this is added to the original code
      )+
      theme(ggh4x.axis.ticks.length.minor = rel(0.5))

It can be used as above. Is this what you were willing for?