coolors

Need a quick color palette for a plot or map in R?

I use coolors.co multiple times a day to find color palettes. It’s well-built, easy-to-use, and produces beautiful palettes. Still, it’s sometimes easier to integrate palette-building within my R workflow, rather than copying and pasting from a web browser; thus, this package was born.

So, why use coolors?

Contains ~2,000 beautiful five-color base palettes, sourced from coolors.co
Allows flexible palette randomization and customization based on your needs
Stores your palette history, so you can mix, match, and compare palettes until you find one you like!

Installation

You can install the development version of this package from GitHub.

# install.packages("devtools")
devtools::install_github("simmwill/coolors")

Basic usage

There are three functions that make this package useful:

palette_coolors() generates a color palette (choosing a random base palette by default)
coolors_history() recalls the current session’s history of generated palettes
shuffle_pal() allows you to change specific colors in a generated palette

Let’s test them out! First, we’ll generate some random color palettes using repeated calls to palette_coolors().

library(coolors)

# generate a few random coolors palettes
palette_coolors()

palette_coolors()

palette_coolors()

Features

Palette history

Since I usually call palette_coolors() several times until I find the palette I like, it’s often helpful to look back on the palettes I’ve called. We can do that using coolors_history(). (History is not preserved between sessions.)

Here, we can see the three palettes above listed in our history by their palette IDs:

# see history, with most recent at top of list
coolors_history()
#> [[1]]
#> Most recent:  
#>           336 
#> 
#> [[2]]
#> [1] 346
#> 
#> [[3]]
#> [1] 450

We can specify a palette by its ID using the palette_num argument. We’ll call palette 450, which was in our history above.

palette_coolors(palette_num = 450)

Shorter/longer palettes

We can also use palette_coolors(n = #) to generate a palette shorter or longer than the default five. If shorter than five, the palette will simply be truncated; if longer, colors will be interpolated.

We’ll use palette 336 as an example:

# 100-color palette (nearly a gradient)
palette_coolors(palette_num = 336, n = 100)

# 2-color palette (chooses the first two colors)
palette_coolors(palette_num = 336, n = 2)

Hex codes

Finally, if we just want the hex codes for a palette, we can use hex = TRUE.

# generate hex codes for palette 336
palette_coolors(palette_num = 336, hex = TRUE)
#> Coolors palette 336
#> [1] "#4f6d7a" "#c0d6df" "#dbe9ee" "#4a6fa5" "#166088"

Custom randomization with `shuffle_pal()`

Let’s say we find a palette that we really love…except for one of the colors. Wouldn’t it be nice to randomize that one color while the others remain the same? We can do this using the shuffle_pal() function!

We’ll use palette 336 from above as an example:

# original palette
palette_coolors(palette_num = 336)

I like this palette, but it’d be nice to replace the third color with another color. To do this, I’ll first call the base palette, 336. Then, I’ll pipe the result to shuffle_pal(), using the freeze argument to keep the colors that I like.

library(magrittr)

palette_coolors(palette_num = 336) %>% 
  shuffle_pal(freeze = c(1:2, 4:5))           # I want to keep colors 1, 2, 4, 5
#> Positions shuffled: 3

Also note that you can specify any number of colors to freeze from 1:1-n, where n is the length of the palette. Here, I call a 7-color palette and shuffle colors 1, 6, and 7 by freeze-ing colors 2 to 5:

# original palette
palette_coolors(palette_num = 123, n = 7)

# shuffled palette
palette_coolors(palette_num = 123, n = 7) %>% 
  shuffle_pal(freeze = c(2:5))
#> Positions shuffled: 1, 6, 7

Iterative randomization

Finally, a more complex use-case that’s surprisingly common the more I use this package. Let’s say from the already-shuffled palette above, I now only want to change the first color.

What if I just change the second line of code to shuffle_pal(freeze = c(2:7)) and rerun it? Unfortunately, this will leave me frustrated. Colors 6 and 7 will be randomized from a different seed than the initial call, and they’ll be changed in addition to color 1.

Instead, we can modify different positions of an already-modified palette in one of two ways: use base R’s automatically-created .Last.value object to modify the last palette, or pipe the hex codes of an already-modified palette directly into shuffle_pal().

.Last.value %>% 
  shuffle_pal(freeze = c(2:7))
#> Positions shuffled: 1

c("#ef626c", "#BB5A57", "#F68A5F", "#FFE066", "#6D9C8C", "#48b8d0", "#9eb3c2") %>% 
  shuffle_pal(freeze = c(2:7))
#> Positions shuffled: 1

We can then keep running either of these chunks of code until we find the right color for position 1.

Using your palette!

So, once we’ve found a palette we like, how can we use it? Perhaps the most straightforward way is to assign the character vector of hex codes produced using palette_coolors(hex = TRUE) to an R object.

Here, we pull hex codes from palette 123 and assign this palette to an object we name pal. We can then use this palette in various R functions, e.g. ggplot2 functions that allow specification using hex codes such as scale_color_manual().

library(ggplot2)
library(palmerpenguins)

pal <- palette_coolors(palette_num = 123, hex = TRUE)[3:5]

ggplot(penguins, aes(x = bill_length_mm, y = flipper_length_mm, color = species)) + 
  geom_point() + 
  scale_color_manual("Species", values = pal) +
  theme_bw() +
  labs(x = "Bill length (mm)",
       y = "Flipper length (mm)",
       title = "Penguin bill vs. flipper length, by species"
  )

Notes

All color palettes courtesy of coolors.co and its creator, Fabrizio Bianchi. This package also draws inspiration from several wonderful color palette packages, including wesanderson.

simmwill/coolors