Skip to contents

Color schemes

Source: vignettes/articles/Color-schemes.Rmd
Color-schemes.Rmd

In this article, we will demonstrate the use of color schemes in tidyplots. We will explore the default color schemes that come with tidyplots and are ready to use for plotting. These include schemes for discrete, continuous and diverging variables. To conclude, we will discuss the creation of custom color schemes from hex values.

Default color schemes

tidyplots comes with a number of default color schemes. Many of them are adapted from the viridisLite and RColorBrewer packages. You access them by loading the the tidyplots library and start typing colors_. The auto-completion will guide you through a selection of discrete, continuous and diverging schemes.

Let’s have a look at the signature scheme of tidyplots colors_discrete_friendly, which was designed to work well for people with color vision deficiency. When running the line colors_discrete_friendly in the console or within a script, a preview of the scheme will be rendered to the Viewer pane in the lower right of the RStudio Desktop interface.

In essence, tidyplots color schemes are just a character vector of hex colors with a special print method that sends a preview to the RStudio viewer pane.

library(tidyplots)
colors_discrete_friendly
colors_discrete_friendly
A tidyplots color scheme with 6 colors.

c(
"#0072B2","#56B4E9","#009E73","#F5C710","#E69F00","#D55E00")

Tip: You can copy individual hex colors directly from the preview to use them in your script.

Discrete

Discrete color schemes are meant for categorical variables. The default schemes in tidyplots consist of 5–7 colors. However, if more categories are present in the plot, tidyplots will automatically fill up the gaps between colors to deliver exactly the number that is required for the plot.

Similarly, when more colors are provided than needed, tidyplots will select the required number of colors by attempting to evenly sample from the supplied color vector.

energy %>% 
  tidyplot(year, power, color = energy_source) %>% 
  add_barstack_absolute()

And here are some alternative color schemes.

energy %>% 
  tidyplot(year, power, color = energy_source) %>% 
  add_barstack_absolute() %>% 
  adjust_colors(colors_discrete_seaside)

energy %>% 
  tidyplot(year, power, color = energy_source) %>% 
  add_barstack_absolute() %>% 
  adjust_colors(colors_discrete_friendly_long)

energy %>% 
  tidyplot(year, power, color = energy_source) %>% 
  add_barstack_absolute() %>% 
  adjust_colors(colors_discrete_apple)

energy %>% 
  tidyplot(year, power, color = energy_source) %>% 
  add_barstack_absolute() %>% 
  adjust_colors(colors_discrete_ibm)

energy %>% 
  tidyplot(year, power, color = energy_source) %>% 
  add_barstack_absolute() %>% 
  adjust_colors(colors_discrete_candy)

Continuous

Continuous color schemes are meant for continuous variables. The default schemes in tidyplots usually consist of 265 colors.

colors_continuous_viridis
colors_continuous_viridis
A tidyplots color scheme with 265 colors, downsampled to 42 colors.

c(
"#440154FF","#460A5DFF","#471264FF","#481B6DFF","#482374FF","#472C7AFF","#46337FFF","#443A83FF","#424186FF","#3F4889FF","#3C508BFF","#39568CFF","#365D8DFF","#33638DFF","#306A8EFF","#2D708EFF","#2B758EFF","#297B8EFF","#26818EFF","#24878EFF","#228D8DFF","#20928CFF","#1F988BFF","#1F9F88FF","#20A486FF","#24AA83FF","#29AF7FFF","#31B57BFF","#3BBB75FF","#45C06FFF","#53C569FF","#5EC962FF","#6ECE58FF","#7BD250FF","#8AD647FF","#9CD93CFF","#AADC32FF","#BDDF26FF","#CCE11EFF","#DEE318FF","#EDE51BFF","#FDE725FF")

Here is a use case for a continuous color scheme.

climate %>% 
  tidyplot(x = month, y = year, color = max_temperature) %>% 
  add_heatmap()

And here are some alternative color schemes.

climate %>% 
  tidyplot(x = month, y = year, color = max_temperature) %>% 
  add_heatmap() %>% 
  adjust_colors(new_colors = colors_continuous_inferno)

climate %>% 
  tidyplot(x = month, y = year, color = max_temperature) %>% 
  add_heatmap() %>% 
  adjust_colors(new_colors = colors_continuous_mako)

climate %>% 
  tidyplot(x = month, y = year, color = max_temperature) %>% 
  add_heatmap() %>% 
  adjust_colors(new_colors = colors_continuous_turbo)

climate %>% 
  tidyplot(x = month, y = year, color = max_temperature) %>% 
  add_heatmap() %>% 
  adjust_colors(new_colors = colors_continuous_rocket)

Diverging

Diverging color schemes are meant for continuous variables that have a central point in the middle. A classical example is the blue–white–red gradient used for gene expression heatmaps.

colors_diverging_blue2red
colors_diverging_blue2red
A tidyplots color scheme with 17 colors.

c(
"#0000FF","#1F1FFF","#3F3FFF","#5F5FFF","#7F7FFF","#9F9FFF","#BFBFFF","#DFDFFF","#FFFFFF","#FFDFDF","#FFBFBF","#FF9F9F","#FF7F7F","#FF5F5F","#FF3F3F","#FF1F1F","#FF0000")

Here is a use case for a diverging color scheme.

gene_expression %>% 
  tidyplot(x = sample, y = external_gene_name, color = expression) %>% 
  add_heatmap(scale = "row") %>%
  sort_y_axis_labels(direction) %>% 
  adjust_size(height = 100)

And here are some alternative color schemes.

gene_expression %>% 
  tidyplot(x = sample, y = external_gene_name, color = expression) %>% 
  add_heatmap(scale = "row") %>%
  sort_y_axis_labels(direction) %>% 
  adjust_size(height = 100) %>% 
  adjust_colors(new_colors = colors_diverging_blue2brown)

gene_expression %>% 
  tidyplot(x = sample, y = external_gene_name, color = expression) %>% 
  add_heatmap(scale = "row") %>%
  sort_y_axis_labels(direction) %>% 
  adjust_size(height = 100) %>% 
  adjust_colors(new_colors = colors_diverging_spectral)

gene_expression %>% 
  tidyplot(x = sample, y = external_gene_name, color = expression) %>% 
  add_heatmap(scale = "row") %>%
  sort_y_axis_labels(direction) %>% 
  adjust_size(height = 100) %>% 
  adjust_colors(new_colors = colors_diverging_icefire)

Custom color schemes

Of course you can also construct custom color schemes using the new_color_scheme() function.

my_colors <- 
  new_color_scheme(c("#ECA669","#E06681","#8087E2","#E2D269"),
                   name = "my_custom_color_scheme")
my_colors
my_custom_color_scheme
A tidyplots color scheme with 4 colors.

c(
"#ECA669","#E06681","#8087E2","#E2D269")

Than you can use your scheme as input to the adjust_colors() function.

energy %>% 
  tidyplot(year, power, color = energy_source) %>% 
  add_barstack_absolute() %>% 
  adjust_colors(new_colors = my_colors)

Besides creating new schemes, you can also subset and concatenate existing schemes in the exact same way you would do with a regular character string.

colors_discrete_metro[2]
Untitled color scheme
A tidyplots color scheme with 1 colors.

c(
"#4FAE62") 

colors_discrete_metro[2:4]
Untitled color scheme
A tidyplots color scheme with 3 colors.

c(
"#4FAE62","#F6C54D","#E37D46") 

c(colors_discrete_metro, colors_discrete_seaside)
Untitled color scheme
A tidyplots color scheme with 10 colors.

c(
"#4DACD6","#4FAE62","#F6C54D","#E37D46","#C02D45","#8ecae6","#219ebc","#023047","#ffb703","#fb8500")

What’s more?

To dive deeper into code-based plotting, here a couple of resources.

tidyplots documentation

Other resources