R Plot Color Combinations that Are Colorblind Accessible

Question

How do I choose 4-8 colors in base R for plots that colorblind people will be able to see?

Below is the base R color pallet. Looking for a solution in BASE R without the use of packages.

Base R Color Palette Guide: http://www.stat.columbia.edu/~tzheng/files/Rcolor.pdf

Color Palettes for Color Blindness: http://mkweb.bcgsc.ca/colorblind/

Specifically how do I manually create accessible colors in BASE R?

e.g. "vermilion", "bluish green", and "reddish purple" as described in the paper figure below.

Answer 1

These are the hex codes for those colors in the image you posted

colorBlindBlack8  <- c("#000000", "#E69F00", "#56B4E9", "#009E73", 
                       "#F0E442", "#0072B2", "#D55E00", "#CC79A7")
pie(rep(1, 8), col = colorBlindBlack8)

colorBlindGrey8   <- c("#999999", "#E69F00", "#56B4E9", "#009E73", 
                       "#F0E442", "#0072B2", "#D55E00", "#CC79A7")
scales::show_col(colorBlindGrey8)

For more than 8 colors, rcartocolor has the Safe palette with 12 colors

safe_colorblind_palette <- c("#88CCEE", "#CC6677", "#DDCC77", "#117733", "#332288", "#AA4499", 
                             "#44AA99", "#999933", "#882255", "#661100", "#6699CC", "#888888")
scales::show_col(safe_colorblind_palette)

If you need sequential or diverging colormaps, check these palettes out

library(rcartocolor)
display_carto_all(colorblind_friendly = TRUE)

^{Created on 2019-07-22 by the reprex package (v0.3.0)}

Answer 2

Okabe-Ito palette

The palette shown in the question is also known as the Okabe-Ito palette as suggested by Okabe & Ito (2008). Since version 4.0.0, base R provides a new palette.colors() where this palette is actually the default:

palette.colors(palette = "Okabe-Ito")
##         black        orange       skyblue   bluishgreen        yellow 
##     "#000000"     "#E69F00"     "#56B4E9"     "#009E73"     "#F0E442" 
##          blue    vermillion reddishpurple          gray 
##     "#0072B2"     "#D55E00"     "#CC79A7"     "#999999" 

Qualitative palettes in base R

Along with this palette, various other qualitative palettes are easily available in base R. Specifically, the new default palette (called "R4") has also been designed to be rather robust under color vision deficiencies. See this blog post for further details:

• Zeileis, Murrell, Maechler, Sarkar (2019). "A New palette() for R." The R Blog.

Sequential and diverging palettes in base R

In addition to the qualitative palettes above, base R also has a new function hcl.colors() since version 3.6.0 that makes many sequential and diverging palettes available that are also robust under color vision deficiencies. It provides approximations (derived using the hue-chroma-luminance color model) to many palettes from ColorBrewer.org, viridis, CARTO colors, Crameri's scientific colors etc. The default is the popular viridis palette. The following blog post provides more details and the paper about the colorspace package explains more related/underlying work.

• Zeileis, Murrell (2019). "HCL-Based Color Palettes in grDevices." The R Blog.
• Zeileis, Fisher, Hornik, Ihaka, McWhite, Murrell, Stauffer, Wilke (2020). "colorspace: A Toolbox for Manipulating and Assessing Colors and Palettes." Journal of Statistical Software.

Answer 3

The color scales in the viridis package are all color blind accessible. https://cran.r-project.org/web/packages/viridis/vignettes/intro-to-viridis.html#the-color-scales

If you need to use colors by name like in the pdf you linked to, match colors from the viridis palettes to the named list of colors. For 4-8, this is pretty easy to do visually.