The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.

Intro to ggtintshade

library(ggplot2)
library(ggtintshade)
library(patchwork)

ggtintshade is an extension to the ggplot2 plotting library that allows for the tint/shade of a color to be mapped to an aesthetic in addition to its hue (color). This allows for visual grouping of similar points in color space while still allowing a color legend to disambiguate them overall. It supports both nested and crossed designs.

Crossed vs nested data

The original inspiration behind ggtintshade came from metabolomics experiments where I wanted to be able to discuss both individual compounds as well as the groups they fell into. A useful visual guide for this is to have all of one compound type be a single color, while individual compounds within that group have different shades.

library(ggtintshade)

metab_data <- data.frame(
  metab = rep(c("Alanine", "Threonine", "Glycine",
                "Glycine betaine", "Proline betaine", "Carnitine",
                "DMSP", "DMS-Ac", "Isethionate"), 3),
  metab_group = rep(rep(c("Amino acid", "Betaine", "Sulfur"), each = 3), 3),
  tripl = rep(c("A", "B", "C"), each = 9),
  area  = runif(27)
)
metab_data$metab <- factor(metab_data$metab, levels = unique(metab_data$metab))

ggplot(metab_data) +
  geom_col_tintshade(aes(x=tripl, y=area, fill=metab_group, tintshade = metab))

This is a nice example of nested data, where each individual entry belongs to a single group. ggtintshade also handles crossed data, where each shade should map into multiple groups. A good example of this is your D&D-style “alignment” chart.

align_data <- data.frame(
  alignment=1:9,
  moral=rep(c("good", "neutral", "evil"), each=3),
  meta=rep(c("lawful", "neutral", "chaotic"), length.out=9)
)
align_data$moral <- factor(align_data$moral, levels=rev(unique(align_data$moral)))
align_data$meta <- factor(align_data$meta, levels=unique(align_data$meta))

ggplot(align_data) +
  geom_raster_tintshade(aes(x=meta, y=moral, fill=meta, tintshade=moral)) +
  scale_fill_manual(breaks = c("lawful", "neutral", "chaotic"), values=c("#cca40a", "#b52060", "#7623b2")) +
  coord_equal()

That last plot also demonstrates how well ggtintshade handles normal ggplot2 behavior. Manually specified colors are handled naturally and the interaction is seamless, but you can also control the degree of lightening/darkening using the expected ggplot2 syntax for the new aesthetic with the associated scale. For example, using the diamonds dataset:

grp <- c(I1 = "I", SI2 = "SI", SI1 = "SI", VS2 = "VS", VS1 = "VS", VVS2 = "VVS", VVS1 = "VVS", IF = "IF")
diamonds$clarity_group <- factor(grp[as.character(diamonds$clarity)], levels = c("I", "SI", "VS", "VVS", "IF"))
mp <- aggregate(price ~ clarity + clarity_group, diamonds, mean)

crossed_gp <- ggplot(mp) +
  geom_col_tintshade(aes(clarity, price, fill = clarity_group, tintshade = clarity)) +
  scale_tintshade_discrete(range = c(0.4, 0.6)) +
  ggtitle("Nested diamonds") +
  theme(axis.text.x = element_text(angle=90, hjust=1, vjust=0.5))
nested_gp <- ggplot(diamonds) +
  geom_bar_tintshade(aes(x=cut, fill = cut, tintshade = clarity), color="black") +
  ggtitle("Crossed diamonds") +
  scale_tintshade_discrete(range = c(0.1, 0.9)) +
  theme(axis.text.x = element_text(angle=90, hjust=1, vjust=0.5))
crossed_gp + nested_gp

Using tintshade instead of alpha

This package is a much better way to map a lightness aesthetic than the advice commonly offered online to use alpha instead. (e.g here, here, here and here) For one, alpha only lightens, not darkens, and creates an unhelpful transparency that must be handled. It also avoids having to calculate an interaction and then set a manual scale for two aesthetics pasted together. Additionally, using an overlapping alpha trace is inefficient and can cause issues if exported to a vectorized format because each point has multiple values.

For example, solving the question raised in Different color shades for faceted grouped bar plots in ggplot2:

temp.data = data.frame (
  Species  = rep(c("A","B"),each=2, times=2),
  Status = rep(c("An","Bac"), times=4),
  Sex = rep(c("Male","Female"), each=4, times=1),
  Proportion = c(6.86, 7.65, 30.13, 35.71, 7.13, 10.33, 29.24, 31.09)
)
init_alpha <- ggplot(temp.data, aes(x = Species, y = Proportion, fill = Status, alpha = Species)) +
  geom_bar(stat='identity', position = position_dodge(width = 0.73), width=.67) +
  facet_grid(Sex ~ .) +
  scale_fill_manual(name = "Status", labels = c("An","Bac"), values = c("#86a681","#0a3e03")) +
  ggtitle("Without ggtintshade")
new_tinted <- ggplot(temp.data, aes(x = Species, y = Proportion, fill = Species, tintshade = Status)) +
  geom_bar_tintshade(stat='identity', position = position_dodge(width = 0.73), width=.67) +
  facet_grid(Sex ~ .) +
  scale_fill_manual(name = "Status", labels = c("An","Bac"), values = c("#cf944c","#0a3e03")) +
  scale_tintshade_discrete(range = c(0.3, 0.7)) +
  ggtitle("With ggtintshade")
init_alpha + new_tinted
#> Warning: Using alpha for a discrete variable is not advised.

Or the question here (grouping multiple gradients using ggplot2) which also demonstrates a continuous tint gradient (and why alpha is a bad idea when points overlap!):

d <- data.frame(
  x=rep(1:20, 5), y=rnorm(100, 5, .2) + rep(1:5, each=20), 
  z=rep(1:20, 5), grp=factor(rep(1:5, each=20))
)

init_alpha <- ggplot(d) +
  geom_path(aes(x, y, color=grp), linewidth=2, lineend=0) +
  geom_path(aes(x, y, group=grp, alpha=z), linewidth=2, lineend=0) +
  ggtitle("Without ggtintshade")
new_tinted <- ggplot(d) + 
  geom_path_tintshade(aes(x, y, color=grp, tintshade=z), linewidth=2, lineend=0) +
  scale_tintshade_continuous(range = c(0.5, 0)) +
  ggtitle("With ggtintshade")
init_alpha + new_tinted

Other examples:

You will have to be careful with the number of colors you’d like to use, just as in the default ggplot2 things. Shading/tinting can add a certain amount of visual chaos that’s sometimes worse than a clean plot.

mpgsub <- head(mpg, 60)
mpgsub$model <- factor(mpgsub$model, levels=unique(mpgsub$model))
subset_plot <- ggplot(mpgsub, aes(displ, hwy, colour = manufacturer, tintshade = model)) +
  geom_point_tintshade(size = 3)

mpg$model <- factor(mpg$model, levels=unique(mpg$model))
all_colors <- ggplot(mpg, aes(displ, hwy, colour = manufacturer, tintshade = model)) +
  geom_point_tintshade(size = 3) +
  guides(tintshade=guide_tintshade(ncol = 1))

subset_plot + all_colors

ggplot(penguins) + 
  geom_point_tintshade(aes(x=bill_len, y=bill_dep, fill=species, tintshade=sex), 
                       pch=21, color="black", size=3)
#> Warning: Removed 2 rows containing missing values or values outside the scale range
#> (`geom_point_tintshade()`).

This last example also shows how NA tintshade values are mapped to the untinted shade, which could cause some confusion. The recommended approach in this case is to ensure that the NA values are also mapped to an additional aesthetic, e.g. shape.

Vignette last built on 2026-07-15

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.