## ----include = FALSE----------------------------------------------------------
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

## -----------------------------------------------------------------------------
library(ggplot2)
library(ggh4x)

# Setting up some random data
n <- 200
df <- data.frame(
  x = c(rpois(n, 25),
        rnbinom(n, 5, 0.2),
        rgamma(n, 30, 1.5),
        rchisq(n, 15)),
  distribution = rep(c("Poisson", "Negative Binomial", 
                       "Gamma", "Chi-squared"), each = n),
  type = rep(c("Discrete", "Continuous"), each = 2 * n)
)

ggplot(df, aes(x, y = after_stat(count),
               fill = distribution, colour = distribution)) +
  geom_histogram(position = "identity", binwidth = 1, 
                 alpha = 0.3, colour = NA) +
  # One type of theoretical densities for discrete distributions with steps
  stat_theodensity(data = ggsubset(type == "Discrete"),
                   distri = "nbinom", geom = "step",
                   position = position_nudge(x = -0.5)) +
  # Another type for the continuous ones with lines
  stat_theodensity(data = ggsubset(type == "Continuous"),
                   distri = "gamma") +
  scale_colour_discrete(aesthetics = c("colour", "fill"), guide = "none") +
  # Have the facet strips span categories
  facet_nested(~ type + distribution, scales = "free_x") +
  # Precisely control aspect ratio of panels
  force_panelsizes(rows = 1.618, cols = 1, respect = TRUE) +
  # Tweak the scales of individual panels
  facetted_pos_scales(list(
    scale_x_continuous(labels = scales::number_format(0.1)),
    # Give the 2nd panel minor ticks
    scale_x_continuous(guide = guide_axis(minor.ticks = TRUE)),
    scale_x_continuous(limits = c(0, 40), oob = scales::oob_keep),
    scale_x_continuous(expand = c(1, 0))
  ))