Intro

We will follow Chapter 12 of Winston Chang’s Graphics Cookbook closely.

It is at https://r-graphics.org/chapter-colors.

Setup

Get the packages we need.

library(tidyverse)

## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.3.6      ✔ purrr   0.3.4 
## ✔ tibble  3.1.8      ✔ dplyr   1.0.10
## ✔ tidyr   1.2.1      ✔ stringr 1.4.1 
## ✔ readr   2.1.2      ✔ forcats 0.5.2 
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

library(socviz)
library(gcookbook)
library(viridis)

## Loading required package: viridisLite

library(RColorBrewer)

A Continuous Color Example

We can use the builtin dataframe mtcars for a simple example.

Do a scatterplot with disp on the x-axis and mpg on the y-axis. Use the aesthetic color to capture the vehicle weight, wt. Save this graphic as an object p and display it.

Solution

p = mtcars %>% ggplot(aes(disp,mpg)) +
  geom_point(aes(color = wt))

p

In this case, points are solid objects and the appropriate aesthetic is “color” instead of fill. The variable wt is continuous, so we get a continuous scale of blues.

Alternatives

Let’s try viridis_c with defaults.

Solution

p + scale_color_viridis_c()

Reverse

Set the direction parameter to -1 in the call to viridis so that darker colors will be associated with higher values.

Solution

p + scale_color_viridis_c(direction = -1)

A Variant

Viridis has different color schemes available. You can get to them with the option parameter in the call to viridis_c(). Potential values are “A” through “H”.

Try a few.

Solution

p + scale_color_viridis_c(direction = -1, option = "C") +
  ggtitle("Option C")

p + scale_color_viridis_c(direction = -1, option = "G") +
  ggtitle("Option G")

Gradients

We can use scale_color_gradient(). We need to pick two colors to represent the low and high values. Here’s an example.

p + scale_color_gradient(low = "white", high = "navy")

Try a few alternatives for named colors using the link to “Named Colors in R” in Moodle.

Solution

p + scale_color_gradient(low = "turquoise", high = "tomato")

p + scale_color_gradient(low = "skyblue", high = "violetred4")

Color Picker

Let’s use the HTML color picker link in R to designate our low and high colors with hex codes.

Solution

p + scale_color_gradient(low = "#FBFE32", high = "#1624E9")

p + scale_color_gradient(low = "#ACFDBF", high = "#B7202B")

A Discrete Variable

For this example we can use a sample of the rows in cdc2. First, we need to get the data.

load("cdc2.Rdata")
scdc2 = cdc2 %>% 
  sample_n(150)
str(scdc2)

## 'data.frame':    150 obs. of  15 variables:
##  $ genhlth    : Factor w/ 5 levels "excellent","very good",..: 1 3 2 4 1 1 5 1 2 3 ...
##  $ exerany    : num  1 1 0 0 1 1 0 1 1 1 ...
##  $ hlthplan   : num  1 1 1 1 1 1 1 1 1 0 ...
##  $ smoke100   : num  1 0 1 1 0 0 1 0 0 1 ...
##  $ height     : num  70 65 72 72 70 67 73 73 69 71 ...
##  $ weight     : int  140 127 200 180 190 165 170 175 210 235 ...
##  $ wtdesire   : int  140 125 185 180 170 155 170 190 210 205 ...
##  $ age        : int  35 32 48 37 55 39 42 35 45 42 ...
##  $ gender     : Factor w/ 2 levels "m","f": 2 2 1 1 1 1 1 1 1 1 ...
##  $ BMI        : num  20.1 21.1 27.1 24.4 27.3 ...
##  $ BMIDes     : num  20.1 20.8 25.1 24.4 24.4 ...
##  $ DesActRatio: num  1 0.984 0.925 1 0.895 ...
##  $ BMICat     : Factor w/ 5 levels "Underweight",..: 2 2 3 2 3 3 2 2 4 4 ...
##  $ BMIDesCat  : Factor w/ 5 levels "Underweight",..: 2 2 3 2 2 2 2 3 4 3 ...
##  $ ageCat     : Factor w/ 4 levels "18-31","32-43",..: 2 2 3 2 3 2 2 2 3 2 ...

Do a scatterplot of height and weight and use ageCat as our categorical variable. First, we’ll use the default colors.

Solution

scdc2 %>% 
  ggplot(aes(height, weight,color = ageCat)) +
  geom_jitter()

## ColorBrewer

Now add a color scale from RColorBrewer. Use the YlOrRd palette.

Solution

scdc2 %>% 
  ggplot(aes(height, weight,color = ageCat)) +
  scale_color_brewer(palette= "YlOrRd") +
  geom_jitter() 

## Two More Examples

Solution

scdc2 %>% 
  ggplot(aes(height, weight,color = ageCat)) +
  scale_color_brewer(palette= "PuBu") +
  geom_jitter() 

scdc2 %>% 
  ggplot(aes(height, weight,color = ageCat)) +
  scale_color_brewer(palette= "Greens") +
  geom_jitter()