Search code examples
rggplot2

Putting label in Histogram in ggplot

I have below histogra musing ggplot

library(ggplot2)
n = 1000
mean = 0
sd = 1
binwidth = 0.3 # passed to geom_histogram and stat_function
set.seed(1)
df <- data.frame(x = rnorm(n, mean, sd))

myplot = ggplot(df, aes(x = x, mean = mean, sd = sd, binwidth = binwidth, n = n)) +
    theme_bw() +
    geom_histogram(binwidth = binwidth, 
        colour = "white", fill = "cornflowerblue", size = 0.1)

Now I want to add one custom label at the top-middl area, but there should be some margin, say 10 points at the top of my label. I tried below

myplot+
geom_label(aes(x = mean((df$x)), y = 100), label = 'Plot with y', label.padding = unit(.7, 'lines'), fill = '#000000', color = '#ffffff', alpha = 0.2)

Is there any way to dynamically (based on data only without any hardcoding) set y = 100 to keep the label at the top with 10 points margin?

Solution

  • One option to place your label at the top of the plot area and having some margin would be to use annotation_custom which requires to add your label via a grob. For the latter I use gridtext::richtext_grob (as grid does not provide a labelGrob whereas ggplot2 does not export it):

    library(ggplot2)
library(gridtext)
library(grid)

myplot +
  annotation_custom(
    grob = gridtext::richtext_grob(
      text = "Plot with y",
      # Place at the top with margin of 5mm
      y = unit(1, "npc") - unit(5, "mm"),
      # Padding around label
      padding = unit(.7, "lines"),
      gp = grid::gpar(col = "#ffffff"),
      box_gp = grid::gpar(fill = "#000000", alpha = .6),
      vjust = 1
    ),
    xmin = mean(df$x),
    xmax = mean(df$x)
  )

    EDIT If you can't install any non-standard package (e.g. gridtext) one option would be to use the non-exported labelGrob from ggplot2 using ::::

    myplot +
  annotation_custom(
    grob = ggplot2:::labelGrob(
      label = "Plot with y",
      # Place at the top with margin of 5mm
      y = unit(1, "npc") - unit(5, "mm"),
      # Padding around label
      padding = unit(rep(.7, 4), "lines"),
      text.gp = grid::gpar(col = "#ffffff"),
      rect.gp = grid::gpar(fill = "black", alpha = .6),
      just = "top"
    ),
    xmin = mean(df$x),
    xmax = mean(df$x)
  )