MattCowgill / ggdirectlabel

ggdirectlabel

The goal of ggdirectlabel is to make it easier to directly label ggplot2 charts rather than using legends.

Installation

You can install the development version of ggdirectlabel from GitHub with:

# install.packages("devtools")
devtools::install_github("MattCowgill/ggdirectlabel")

library(ggdirectlabel)
library(ggplot2)
library(magrittr)

Using geom_richlegend()

Here’s a standard ggplot2 scatterplot:

base_scatter <- mtcars |> 
  ggplot(aes(x = wt, y = mpg, col = factor(cyl))) +
  geom_point()

base_scatter

This is fine! But sometimes you might like the legend levels (4, 6, and 8 in this example) to be coloured according to the levels in the data. That’s where geom_richlegend() comes in:

base_scatter +
  geom_richlegend(aes(label = cyl)) +
  theme(legend.position = "none")

You can move the ‘rich legend’ around:

base_scatter +
  geom_richlegend(aes(label = cyl),
                  legend.position = "bottomleft",
                  vjust = 0,
                  hjust = 0) +
  theme(legend.position = "none")

geom_richlegend() respects facets - it’ll place a little legend annotation for each level of the data that appears in that panel:

base_scatter +
  geom_richlegend(aes(label = paste0(cyl, " cylinders"))) +
  facet_wrap(~cyl)

Using geom_linepoint()

Without ggirectlabel, we might do something like:

ggplot2::economics_long %>%
  ggplot(aes(x = date, y = value, col = variable)) +
  geom_line() +
  geom_point(data = ~dplyr::filter(., date == max(date)),
             fill = "white",
             shape = 21,
             size = 2.5,
             stroke = 1.25)

This is fine! But this is a more straightforward way to achieve the same thing:

ggplot2::economics_long %>%
  ggplot(aes(x = date, y = value, col = variable)) +
  geom_linepoint()

Using scale_x_date_rightalign()

In time series line charts, it’s often important to make clear the date of your most recent observation. The scale_x_date_rightalign() function aligns the breaks of your x-axis so that the most recent observation is included in the breaks.

ggplot2::economics_long %>%
  ggplot(aes(x = date, y = value, col = variable)) +
  geom_linepoint() +
  scale_x_date_rightalign()

Using geom_finallabel()

In time series line charts, you may wish to label the final point in the series. The geom_finallabel() function makes that easy.

ggplot2::economics_long %>%
  ggplot(aes(x = date, y = value, col = variable)) +
  geom_linepoint() +
  geom_finallabel(aes(label = round(value, 0))) +
  scale_x_date_rightalign(expand = expansion(c(0, 0.15))) +
  theme(legend.position = "none")