Grouped overlapping histograms

[stefaneng]

One feature from ggplot2 that is not obvious how to do in BoutrosLab.plotting.general is to group a histogram by some categorical variable and overlay on the same plot. In ggplot2 this would correspond to:

set.seed(13);
# Sample size
N <- 1000;
# Generate some data in two groups
data <- rbind.data.frame(
  cbind.data.frame(value = rnorm(N), group = 1),
  cbind.data.frame(value = rnorm(N, mean = 1.5), group = 2)
  );

data$group <- as.factor(data$group)

library(ggplot2)
ggplot(data) +
    geom_histogram(aes(x = value, fill = group), alpha = 0.5, position = 'identity')
#> `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

^{Created on 2023-01-12 with reprex v2.0.2}

[stefaneng]

A solution in BoutrosLab.plotting.general is

set.seed(13);
# Sample size
N <- 1000;
# Generate some data in two groups
data <- rbind.data.frame(
  cbind.data.frame(value = rnorm(N), group = 1),
  cbind.data.frame(value = rnorm(N, mean = 1.5), group = 2)
  );
null.sim <- rnorm(N);
alt.sim <- rnorm(N, mean = 2);
group1.col <- '#E69F00';
group2.col <- '#56B4E9';
histogram1 <- create.histogram(
  x = data$value[data$group == 2],
  type = 'count',
  ylab.label = 'Count',
  xlab.label = 'value',
  col = scales::alpha(group1.col, alpha = 0.7),
  xlimits = c(-5, 6),
  # Add key to upper left corner
  key = list(
    corner = c(0,.95),
    rect = list(
      col = c(group1.col, group2.col)
      ),
    text = list(c('Group 1', 'Group 2'))
    )
  );
histogram2 <- create.histogram(
  x = data$value[data$group == 1],
  type = 'count',
  col = scales::alpha(group2.col, alpha = 0.7)
  );
# Simply add the two histograms together
histogram1 + histogram2