ggCheysson

Version 1.2.0; documentation built 2026-11-04

The ggCheysson package brings the graphical styles of the Albums de Statistique Graphique to R and ggplot2.

The Albums were produced by the Ministry of Public Works in France under the direction of Émile Cheysson from 1879-1897. They represent the “pinnacle of the Golden Age of Statistical Graphics” (Friendly, 2008) for their innovation in visualization techniques, graphic design and beauty.

The package is based on work by:

David Rumsey Map Collection: Complete digitized Albums de Statistique Graphique View collection
RJ Andrews: Classic Map Color Design | GitHub
Tom Shanley: Observable implementation Cheysson Color Palettes
Kenneth Fields: Hand-drawn font family creation The style of Émile Cheysson

🏛️ Historical Context

The Albums de Statistique Graphique were published annually by France’s Ministry of Public Works, showcasing infrastructure statistics through innovative visualizations. Under Émile Cheysson’s direction, these albums combined:

Sophisticated use of color to represent quantitative variables
Hatching patterns to differentiate categories and show density
Hand-lettered text with distinctive character
Clear hierarchical organization of information

This package preserves these design elements for modern statistical graphics. Perhaps more interesting is the historical motivation that led to this project by me and the others mentioned above. Émile Cheysson and his draftspeople at the Ministry of Public Works drew all these plates by hand.

Did they have an “in-house” style guide?
How would graphic designers of today approach this task?

📂 Installation

Install the development version from GitHub or R-universe:

# install.packages("remotes")
remotes::install_github("friendly/ggCheysson")

# or, from R-universe
install.packages("ggCheysson", repos = c("https://friendly.r-universe.dev"))

Suggested Dependencies

For full functionality, install these packages:

install.packages(c("ggpattern", "systemfonts"))

✨ Features

This package provides a complete aesthetic system for creating visualizations in Cheysson’s distinctive style:

🎨 Color Palettes

25 authentic color palettes extracted from the original Albums
Sequential, diverging, grouped, and categorical palette types
Named by album year and plate number (e.g., 1880_07, 1881_12)
Compatible with standard ggplot2 color scales

📐 Hatching Patterns

134 pattern specifications including solid fills, stripes, and crosshatching
Line angles (0°, 45°, 90°, 135°) matching historical diagrams
Variable densities and line widths
Full integration with ggpattern

✍️ Authentic Fonts

5 hand-drawn font families replicating Cheysson’s lettering style
Regular, Italic, Sans Caps, Outline Caps, and Title variants
Automatic loading and integration with ggplot2 themes
Created by Kenneth Fields for historical accuracy

🎭 Complete Themes

theme_cheysson() - Full period-appropriate theme
theme_cheysson_minimal() - Minimal grid variant
theme_cheysson_map() - Optimized for cartographic work

These were derived by RJ Andrews from a collection of 25 thematic maps across the span of years in which the Albums were produced, shown below.

Grid of original historical maps from the Albums de Statistique Graphique, showing the variety of colors and hatching patterns used

From these, he abstracted the following combinations of color and pattern he thought characterized these maps:

Swatches of the color and pattern combinations abstracted from the maps above

New here

In the development version (1.2.0), scale_cheysson() (experimental) applies Cheysson’s actual designed color-pattern pairings in a single call. Each element of a palette is a swatch, a ggpattern pattern that draws its fill color, hatch type, line colors, angle and line spacing together, so a plot needs only aes(pattern = x) and one scale (see the examples below, and vignette("combining-colors-patterns")). The individual pattern scales, scale_pattern_type_cheysson() and the rest, remain for control over each property.

🚀 Quick Start

Here are a few examples to get you started.

Basic Color Palette

Use the color palette of the 1881 Album, plate 4

library(ggplot2)
library(ggCheysson)

ggplot(iris, aes(Sepal.Length, Sepal.Width, color = Species)) +
  geom_point(size = 3) +
  scale_color_cheysson("1881_22") +
  labs(title = "Iris Dataset") +
  theme_minimal()

Scatterplot of iris sepal width against sepal length, with points colored by species in the muted rose, pale yellow and sage green of palette 1881_22

With Fonts and Theme

# Load Cheysson fonts (once per session)
load_cheysson_fonts(method = "showtext")
# Enable showtext for rendering
showtext::showtext_auto()

p1 <- ggplot(mtcars, aes(wt, mpg, color = factor(cyl))) +
  geom_point(size = 3) +
  scale_color_cheysson("1883_31") +
  labs(
    title = "Automobile Efficiency",
    subtitle = "Weight vs Fuel Economy",
    x = "Weight (1000 lbs)",
    y = "Miles per Gallon"
  )

p1 + theme_cheysson()

Scatterplot of mpg against car weight from mtcars, colored by number of cylinders in slate blue, brown and orange from palette 1883_31, drawn with theme_cheysson: a bold outlined capital title, a serif subtitle and small-capital axis titles

Scaling with `base_size`

Like other ggplot2 themes, theme_cheysson() takes a base_size for larger output (slides, posters). Title and axis-title text (set in the decorative CheyssonTitle/CheyssonSansCaps fonts) is corrected to scale at the same rate as the rest of the theme, rather than reading undersized as base_size grows:

p1 + theme_cheysson(base_size = 14)

The same mtcars scatterplot with theme_cheysson(base_size = 14): title, subtitle, axis and legend text are all larger, in proportion

p1 + theme_cheysson(base_size = 16)

The same mtcars scatterplot with theme_cheysson(base_size = 16): the text is larger still, with the title fonts scaled in step with the rest

Complete Cheysson Aesthetic (Colors + Patterns + Fonts)

library(ggpattern)

data <- data.frame(
  category = LETTERS[1:4],
  value = c(15, 23, 18, 20)
)

# One scale applies the palette's colors, hatch types, angles and spacings together
ggplot(data, aes(category, value, pattern = category)) +
  geom_col_pattern(color = "black") +
  scale_cheysson("1883_30") +
  labs(
    title = "Statistical Comparison",
    x = "Category",
    y = "Value"
  ) +
  theme_cheysson() +
  theme(legend.position = "none")

Bar chart of four categories, A to D, in theme_cheysson, each bar hatched with a swatch from palette 1883_30: red diagonal stripes, blue diagonal stripes, red and blue crosshatching, and black diagonal stripes

🌈 Available Palettes

cheysson_palettes provides 25 named color palettes, each with its type (sequential, diverging, grouped, or category), the album year and plate number it was extracted from, and its hex color codes.

Click to show/hide the list of all palettes

library(ggCheysson)

# List all palettes
list_cheysson_pals()
#>       name       type album plate n_colors
#> 1  1880_07    grouped  1880     7        5
#> 2  1880_21   category  1880    21        7
#> 3  1881_12 sequential  1881    12        1
#> 4  1881_14    grouped  1881    14        2
#> 5  1881_22   category  1881    22        4
#> 6  1881_30    grouped  1881    30        5
#> 7  1882_18    grouped  1882    18        2
#> 8  1883_21  diverging  1883    21        3
#> 9  1883_30   category  1883    30        4
#> 10 1883_31  diverging  1883    31        2
#> 11 1886_11    grouped  1886    11        4
#> 12 1886_17    grouped  1886    17        3
#> 13 1886_18    grouped  1886    18        3
#> 14 1886_24    grouped  1886    24        3
#> 15 1886_26 sequential  1886    26        2
#> 16 1886_28   category  1886    28        3
#> 17 1887_22    grouped  1887    22        2
#> 18 1888_27 sequential  1888    27        1
#> 19 1891_14    grouped  1891    14        2
#> 20 1891_19 sequential  1891    19        1
#> 21 1891_25 sequential  1891    25        2
#> 22 1895_16 sequential  1895    16        3
#> 23 1900_28 sequential  1900    28        2
#> 24 1906_06   category  1906     6        6
#> 25 1906_50   category  1906    50        4

# List by type
list_cheysson_pals("sequential")
#>      name       type album plate n_colors
#> 1 1881_12 sequential  1881    12        1
#> 2 1886_26 sequential  1886    26        2
#> 3 1888_27 sequential  1888    27        1
#> 4 1891_19 sequential  1891    19        1
#> 5 1891_25 sequential  1891    25        2
#> 6 1895_16 sequential  1895    16        3
#> 7 1900_28 sequential  1900    28        2

# View palette colors
cheysson_pal("1880_21")
#> [1] "#d9636c" "#869e80" "#dec367" "#85aab1" "#aea9a4" "#ed8238" "#ab90a4"

Visualizing Palettes

Use show_palette() to display a palette with color swatches and hex codes:

# Display a single palette with metadata
show_palette("1895_16")

show_palette() display of palette 1895_16: a header giving its type (sequential), album 1895 and plate 16, above three color rectangles, sage grey, slate blue and near-black, each labeled with its hex code

# Display multiple palettes at once
show_palettes(c("1880_21", "1906_06", "1881_30"))

show_palettes() display of three palettes stacked vertically, each with a header and labeled color rectangles: category palette 1880_21 with 7 colors, category palette 1906_06 with 6 colors, and grouped palette 1881_30 with 5 colors

# One palette of each type in a 2x2 grid: category, sequential, diverging, grouped
show_palettes(c("1880_21", "1895_16", "1883_21", "1881_30"), ncol = 2)

show_palettes() display in a 2 by 2 grid, one palette of each type: category 1880_21 (7 colors), sequential 1895_16 (3 colors), diverging 1883_21 (red, cream, yellow) and grouped 1881_30 (5 colors)

Some palettes, such as the sequential 1881_12, have only one color: Cheysson made the steps with hatching of increasing density rather than with shades. Their swatches here show a single color; use cheysson_pattern("1881_12") to get the full palette (see Pattern Support, below).

# Display all palettes of a specific type
show_palettes("category")

Palette types:

Sequential (7 palettes): For ordered quantitative data
Diverging (2 palettes): For data with neutral midpoint
Grouped (10 palettes): For comparing related groups
Category (6 palettes): For categorical data

📐 Pattern Support

With ggpattern, recreate the distinctive hatching styles - each palette pairs colors with historically accurate patterns (solid fills, stripes, crosshatching):

library(ggpattern)

# List available pattern palettes
head(list_cheysson_patterns())
#>      name       type album plate n_patterns
#> 1 1880_07    grouped  1880     7          7
#> 2 1880_21   category  1880    21          7
#> 3 1881_12 sequential  1881    12          3
#> 4 1881_14    grouped  1881    14          4
#> 5 1881_22   category  1881    22          4
#> 6 1881_30    grouped  1881    30          8

# Each element of a palette is a bundle of properties
patterns <- cheysson_pattern("1886_24")
str(patterns[[2]])
#> List of 7
#>  $ type             : chr "crosshatch"
#>  $ fill             : chr "transparent"
#>  $ pattern_fill     : chr "#b61002"
#>  $ pattern_color    : chr "#b61002"
#>  $ pattern_density  : num 0.974
#>  $ pattern_spacing  : num 0.0263
#>  $ pattern_linewidth: num 2

# Display the palette's elements as swatches
n <- length(patterns)
ggplot(data.frame(i = factor(seq_len(n))), aes(i, 1, pattern = i)) +
  geom_col_pattern(color = "black", width = 0.95) +
  scale_cheysson("1886_24") +
  theme_void() +
  theme(legend.position = "none")

The eight swatches of palette 1886_24 side by side: solid red, red crosshatching, red diagonal stripes, solid blue, blue crosshatching, solid black, black crosshatching and black diagonal stripes

Use scale_cheysson() in a plot to give each category a swatch:

data <- data.frame(category = LETTERS[1:4], value = c(15, 23, 18, 20))

ggplot(data, aes(category, value, pattern = category)) +
  geom_col_pattern(color = "black") +
  scale_cheysson("1886_24") +
  labs(title = "Statistical Comparison", x = "Category", y = "Value") +
  theme_minimal()

Bar chart of four categories using scale_cheysson("1886_24"): A is solid red, B red crosshatching, C red diagonal stripes and D solid blue, with a legend whose keys show the same swatches

All the swatches

show_swatches() draws every swatch, one row per palette, grouped by palette type. The swatch in row 1886_24, column 2 is "1886_24-2", and scale_cheysson("1886_24") uses that row from left to right:

show_swatches()

Atlas of all 134 Cheysson swatches: one row per palette, grouped into panels by palette type (sequential, diverging, grouped, category), with columns numbered by element; each cell shows a solid fill, stripes or crosshatching

For control over each property separately, use the individual pattern scales (scale_fill_cheysson_pattern(), scale_pattern_type_cheysson(), scale_pattern_fill_cheysson(), …), mapping each aesthetic yourself.

✍️ Font Families

Five Cheysson font families are included:

Family	Description	Use
`Cheysson`	Regular serif	Body text, labels
`CheyssonItalic`	Italic variant	Emphasis
`CheyssonSansCaps`	Sans capitals	Axis titles
`CheyssonOutlineCaps`	Outlined caps	Decorative titles
`CheyssonTitle`	Display font	Main titles

Two of these, CheyssonTitle and CheyssonOutlineCaps are designed to be decorative, mainly used in titles

Sample lettering in the two Cheysson font families: outline capitals, and title

Here are the others:

Sample lettering in the three Cheysson font families: regular, italic, sans capitals

To use these:

# Load fonts
load_cheysson_fonts(method = "showtext")
showtext::showtext_auto()

# View font metadata
cheysson_fonts

# Use specific fonts
theme(
  plot.title = element_text(family = "CheyssonTitle"),
  axis.title = element_text(family = "CheyssonSansCaps"),
  axis.text = element_text(family = "Cheysson")
)

📦 Package Contents

Data

cheysson_palettes - Color palette specifications (25 palettes)
cheysson_patterns - Pattern/hatching specifications (134 patterns)
cheysson_fonts - Font family metadata (5 fonts)
albumImages - Metadata linking palettes to original album plates

Color Functions

cheysson_pal() - Get colors from a palette
scale_color_cheysson() / scale_fill_cheysson() - ggplot2 color scales
list_cheysson_pals() - List available palettes
show_palette() - Display a single palette with color swatches and hex codes
show_palettes() - Display multiple palettes for comparison

Pattern Functions

scale_cheysson() - Apply a color-and-pattern palette in one step (experimental)
cheysson_pattern() - Get pattern specifications
scale_pattern_*_cheysson() - ggpattern scales for single properties: fills, types, angles, densities
list_cheysson_patterns() - List available pattern palettes
show_swatches() - Display all the swatches, or those of chosen palettes or types

Font Functions

load_cheysson_fonts() - Load font families
cheysson_font() - Get font family names
cheysson_fonts_available() - Check font availability

Themes

theme_cheysson() - Complete Cheysson theme
theme_cheysson_minimal() - Minimal variant
theme_cheysson_map() - For maps

🙏 Sources and Attribution

Color Palettes and Patterns

David Rumsey Map Collection: Complete digitized Albums de Statistique Graphique View collection
RJ Andrews: SVG pattern extraction and digitization Classic Map Color Design | GitHub
Tom Shanley: Observable implementation Cheysson Color Palettes

Fonts

Kenneth Fields: Hand-drawn font family creation The style of Émile Cheysson

🖼️ Gallery

Original maps from the Albums de Statistique Graphique showing the variety of colors and patterns

Original maps from the Albums showing the variety of colors and patterns

Grid of 25 color palettes extracted by RJ Andrews from the Albums de Statistique Graphique maps

Extracted color palettes by RJ Andrews

🚧 Development Status

This package is under active development. Current features are stable and tested, but the API may evolve. Feedback and contributions are welcome!

ggpattern - Pattern fills for ggplot2
ggthemes - Additional themes for ggplot2
systemfonts - Font handling for R

⚖️ License

GPL (>= 3)

📖 Citation

To cite ggCheysson, please use:

citation("ggCheysson")
#> To cite package 'ggCheysson' in publications use:
#> 
#>   Friendly M (2026). _ggCheysson: Graphic Styles of Emile Cheysson for
#>   'ggplot2'_. R package version 1.2.0,
#>   <https://github.com/friendly/ggCheysson>.
#> 
#> A BibTeX entry for LaTeX users is
#> 
#>   @Manual{,
#>     title = {ggCheysson: Graphic Styles of Emile Cheysson for 'ggplot2'},
#>     author = {Michael Friendly},
#>     year = {2026},
#>     note = {R package version 1.2.0},
#>     url = {https://github.com/friendly/ggCheysson},
#>   }

To cite the original Albums de Statistique Graphique:

France. Ministère des travaux publics. Album de statistique graphique de [year]. Paris: Imprimerie nationale, [1879-1897].

📚 References

Friendly, M. (2008). The Golden Age of Statistical Graphics. Statistical Science, 23(4), 502–535. https://doi.org/10.1214/08-STS268

mirror server hosted at Truenetwork, Russian Federation.