ggpattern

ggpattern provides custom ggplot2 geoms which support filled areas with geometric and image-based patterns.

Reading the articles/vignettes on the package website is probably the best way to get started.

Feature Summary

• Custom versions of (almost) all the geoms from ggplot2 which have a region which can be filled.
• A suite of aesthetics for controlling the pattern appearance (e.g. pattern_alpha)
• The ability to include user-defined patterns

Installation

You can install the development version from GitHub with the instructions below.

Note: there are a lot of dependencies. Please ensure these are installed first before attempting to install ggpattern. By doing so, this will make any issues during installation clearer, leading to more meaningful bug reporting.

install.packages('ggplot2')
install.packages('scales')
install.packages('grid')
install.packages('glue')
install.packages('rlang')
install.packages('tibble')
install.packages('gridGeometry')
install.packages('sf')
install.packages('png')
install.packages('magick')
install.packages('memoise')

# install.packages("remotes")
remotes::install_github("coolbutuseless/ggpattern")

Quickstart

1. Take an existing plot which contains a geom with a fillable area e.g geom_col().
2. Use the {ggpattern} version of the geom e.g. ggpattern::geom_col_pattern() instead of ggplot2::geom_col()
3. Set the aesthetic pattern to your choice of pattern e.g pattern = 'stripe', and set other options using pattern_* aesthetics

df <- data.frame(level = c("a", "b", "c", 'd'), outcome = c(2.3, 1.9, 3.2, 1))

ggplot(df) +
  geom_col_pattern(
    aes(level, outcome, pattern_fill = level), 
    pattern = 'stripe',
    fill    = 'white',
    colour  = 'black'
  ) +
  theme_bw(18) +
  theme(legend.position = 'none') + 
  labs(
    title    = "ggpattern::geom_pattern_col()",
    subtitle = "pattern = 'stripe'"
  ) +
  coord_fixed(ratio = 1/2)

Gallery

Feature Details

Available Geoms

ggpattern includes versions of (nearly) all geoms from ggplot2 which could plausiblly support being filled with a pattern.

See the vignette galleries for examples of all the available geoms filled with geometry-based patterns and image-based/array-based patterns.

New aesthetics

To control pattern appearance, a raft of new aesthetics have been added. e.g.  pattern_alpha, pattern_filename, pattern_density.

There are also scale functions to control each of these new aesthetics e.g.  scale_pattern_alpha_discrete.

Not all aesthetics apply to all patterns. See the individual pattern vignettes for which aesthetics it uses, or see the first vignette on developing user-defined patterns for a table of aesthetic use by pattern, or see the individual vignettes for each pattern.

User-Defined Patterns

Users can write their own pattern functions and ask ggpattern to use them, without having to include the pattern in the package.

See the vignettes on developing patterns ( 1 2, 3 ) for how to do this, and see the vignettes on experimental patterns to see this in action ( Point filling, Hex pattern, Ambient Noise ).

Vignettes

General examples

• geom gallery (geometry-based patterns) Examples of every geom filled with the geometry-based patterns (i.e. ‘stripe’, ‘crosshatch’, ‘circle’)
• geom gallery (array-based patterns) Examples of every geom filled with the array-based patterns (i.e. ‘image’, ‘magick’, ‘gradient’, ‘plasma’, ‘placeholder’)

Exploration of pattern parameters and appearance

• Geometry-based patterns
  • Common aesthetics for geometry-based patterns
  • stripes
  • crosshatch
  • circles
• Array-based patterns
  • image
  • placeholder
  • gradient
  • plasma
  • magick

Developing your own pattern

• Devloping Patterns 1 - Pattern Overview
• Devloping Patterns 2 - Geometry-based pattern
• Devloping Patterns 3 - Array-based pattern

Experimental patterns

These are patterns that aren’t quite ready for prime-time. Feel free to steal the code and extend to suit your needs.

• Point filling
• Hex pattern
• Ambient Noise

Other examples

• gganimate

Limitations

• Nearly always need to use coord_fixed() to ensure the aspect ratio is calculated correctly. Use pattern_aspect_ratio to override the internal calculation, of for occasions where you can’t use coord_fixed() because a different coord_*() is used.
• Legend rendering for patterns is still not great.
  • Use pattern_key_scale_factor to adjust legend appearance.
• The Rstudio output device can be quite slow for plots with lots of patterns. It is often faster to save directly to PNG or PDF and view that.
• Self intersecting geometry can be an issue.
• Non-linear coordinate systems have not been tested.
• Polygons with holes are not supported

ToDo

• Possibly add geoms from third-party sources e.g.
  • geom_circle() and geom_voronoi() from ggforce
• A vignette on how array-based patterns are implemented.
• A vignette on why aspect ratio is hard to get right.