liupeng23/gguidance

A 'ggplot2' extension that focusses on expanding the plotter's arsenal of guides, such as axes, legends and colour bars. It brings various improvements to existing guides by re-implementation, as well as adding novel guides.

gguidance

Warning You’re looking at an experimental branch that explores guide extensions with an upcoming ggproto overhaul in ggplot2.

The goal of gguidance is to provide additional guides to the ggplot2 ecosystem.

Please note that this repo is still being worked on and, while probably usable, isn’t finished.

Installation

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

# install.package("pak")
pak::pak("tidyverse/ggplot2") # Installs development version of ggplot2
pak::pak("teunbrand/gguidance@main") # Installs this package

Overview

A table of contents of this package’s utilities:

1. Legends
   1. Cross legend
   2. String legend
2. Colour bars
   1. Capped colour bar
   2. Violin colour bar
   3. Histogram colour bar
3. Axes
   1. Subtitles
   2. Minor ticks
   3. Capping
   4. Ranges
   5. Table
4. Grids
   1. Regular
   2. Plus
   3. Zebra

Examples

Let’s first set up a basic plot to experiment with

library(gguidance)
#> Loading required package: ggplot2

p <- ggplot(mpg, aes(displ, hwy)) +
  geom_point() +
  labs(
    x = "Engine displacement",
    y = "Highway miles per gallon"
  ) +
  theme(axis.line = element_line())

Legends

Cross legend

A ‘cross legend’ can show two variables in a single legend.

p + aes(colour = paste(cyl, year)) +
  guides(colour = "legend_cross")

String legend

A string legend doesn’t display keys, but colours the labels.

p + aes(colour = class) +
  guides(colour = "legend_string")

Colour bars

Capped colour bar

A capped colour bar:

p + aes(colour = cty) +
  scale_colour_viridis_c(guide = "colourbar_cap")

Violin colour bar

Using a violin as a colour guide:

p + aes(colour = cty) +
  scale_colour_viridis_c(guide = guide_colour_violin(density = mpg$cty))

Histogram colour bar

Using a histogram as a colour guide:

p + aes(colour = cty) +
  scale_colour_viridis_c(
    guide = guide_colour_histogram(hist = hist(mpg$cty, breaks = 10))
  )

Axes

Subtitles

Using subtitles.

p + guides(x = guide_axis_extend(subtitle = c("Less", "More")))

Minor ticks

Using minor ticks.

p + guides(x = "axis_minor")

Capping

Using capped lines.

p + guides(x = "axis_cap")

Ranges

With bracketed ranges.

boxplot <- ggplot(mpg, aes(class, displ)) +
  geom_boxplot() +
  labs(
    x = "Type of car",
    y = "Engine displacement"
  )

boxplot +
  guides(x = guide_axis_nested(
    range_start = c(0.5, 3.5),
    range_end   = c(4.5, 6.5),
    range_name  = c("First range", "Second range"),
    bracket     = "square" 
  ))

Table

Using a table as an axis guide.

# Creating summary table
my_table <- lapply(split(mpg[, c("displ", "cty", "hwy")], mpg$class), colMeans)
my_table <- as.data.frame(do.call(rbind, my_table))
my_table[] <- lapply(my_table, scales::number, accuracy = 0.01)
my_table$class <- rownames(my_table)

# Use summary table as axis guide
boxplot +
  guides(x = guide_axis_table(table = my_table, key_col = class))

Grids

Regular

Like the vanilla panel grid, but with more options for where breaks appear. Here, between categories.

boxplot +
  coord_guided(guide_grid(x_breaks = breaks_between()))

Plus

Subtler grid lines by only drawing the intersections, making a ‘plus’ symbol.

p + coord_guided("grid_plus")

Zebra

Alternating stripes between breaks.

boxplot + coord_guided("grid_zebra") +
  theme(panel.grid.major.x = element_line(alpha("grey50", 0.1)))