ggplot2: how to color specific bins in histogram based on bin ranges

Question

I am plotting a histogram with ggplot2 and trying to figure out how to color specific bins in another color than the others. The bins I want to color are defined by their bin edges / ranges.

Similar questions I found were asking for conditional coloring based on the original values, not the bin ranges, either for a specific value or using a threshold.

Example:

dt <- data.table(x = runif(10000))

ggplot(dt, aes(x)) + 
  geom_histogram(binwidth = 0.01, boundary = 0, closed = "left", 
                 col = "darkgreen", fill = "darkgreen", alpha = 0.5, size = 0.1) +
  scale_x_continuous(breaks = seq(0, 1, 0.1))

which gives me this plot:

I defined the leftmost bin to be [0, 0.01), from there the others are calculated.

Now I want to color the following bins differently: [0, 0.01), [0.1, 0.11), [0.2, 0.21) ..., i.e. the bins starting at

> seq(0, 1, 0.1)
[1] 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

How can I do this?

---

EDIT: this is my desired plot:

Answer 1

davidnortes' answer shows colors well, here's an option if you only want to highlight some bins. I'll start with cut to pre-identify some bins (this needs to be kept in-sync with your binwidth= and other options to geom_histogram), and then a simple logical to determine which ones to highlight.

library(dplyr)
dt %>%
  mutate(
    grp = cut(x, seq(0, 1, by = 0.01), labels = FALSE, include.lowest = TRUE),
    is6 = between(grp, 60, 69)
  ) %>%
  ggplot(aes(x, fill = is6)) +
    geom_histogram(binwidth = 0.01, boundary = 0, closed = "left", 
                   col = "darkgreen", alpha = 0.5, size = 0.1) +
    scale_x_continuous(breaks = seq(0, 1, 0.1))

Options:

• to remove the is6 legend, add + scale_fill_discrete(guide = "none")

• if you want multiple distinct bands, perhaps case_when can help, noting that is6 does not need to be logical:

  dt %>%
    mutate(
      grp = cut(x, seq(0, 1, by = 0.01), labels = FALSE, include.lowest = TRUE), 
      highlight = case_when(
        between(grp, 60, 69) ~ "A",
        between(grp, 20, 25) ~ "B",
        TRUE ~ "C")
    ) %>%
    ggplot(aes(x, fill = highlight)) +
      geom_histogram(binwidth = 0.01, boundary = 0, closed = "left", 
                     col = "darkgreen", alpha = 0.5, size = 0.1) +
      scale_x_continuous(breaks = seq(0, 1, 0.1)) 
  

  The scale_fill_discrete works for this, too.

• you may want specific colors for each group of highlight or such, use scale_fill_manual.

---

Edit:

Here's your image, colors notwithstanding:

dt %>%
  mutate(
    grp = (x %% 0.1 < 0.01)
  ) %>%
  ggplot(aes(x, fill = grp)) +
    geom_histogram(binwidth = 0.01, boundary = 0, closed = "left", 
                   col = "darkgreen", alpha = 0.5, size = 0.1) +
    scale_x_continuous(breaks = seq(0, 1, 0.1))

Answer 2

If you want to create ranges of values along your variable X and color them differently, you can use the cut function:

cut divides the range of x into intervals and codes the values in x according to which interval they fall. The leftmost interval corresponds to level one, the next leftmost to level two and so on.

So, tweaking a little your code, you would have:

#Grouping variable 'x' in dt according sequence 'seq(0, 1, 0.1)'    
dt$breaks <- cut(dt$x, breaks = seq(0, 1, 0.1))

#Plotting
ggplot(dt, aes(x, col = breaks, fill = breaks)) + 
  geom_histogram(binwidth = 0.01, boundary = 0, closed = "left", alpha = 0.5, size = 0.1) +
  scale_x_continuous(breaks = seq(0, 1, 0.1))