Introduction
Mathematical expressions are super common in figures: stats results or text labels that involve italic fonts, subscripts and superscripts, Greek letters, and math symbols, just to name a few. In this quick post, I’m going to show you three ways to add math expressions to ggplots. Read on!
Mathematical expressions in ggplots
(1) Math expressions using expression()
We can use the function expression() in base R to create a wide variety of math expressions: italic and bold fonts, superscripts and subscripts, Greek letters, hat and bar symbols, fractions, integrals and derivatives, etc. Here are some commonly used ones (see ?plotmath for a complete list of math expressions!):
library(tidyverse)
### Common math expressions and symbols
ggplot() +
annotate(geom = "text", x = 1, y = 1, label = expression(Italic~and~bold~font:~This~is~italic(italic~R)~and~bold(bold~R)~and~bolditalic(bold~italic~R)), size = 5, hjust = 0) +
annotate(geom = "text", x = 1, y = 0.8, label = expression(Superscript~and~subscript:~italic(F[1*","*20]) == 5.2*";"~italic(R^2) == 0.8), size = 5, hjust = 0) +
annotate(geom = "text", x = 1, y = 0.6, label = expression(Greek~letter:~italic(alpha) == 0.05*";"~italic(chi^2) == 3.5), size = 5, hjust = 0) +
annotate(geom = "text", x = 1, y = 0.4, label = expression(Math~annotation:~y == frac(1, sigma*sqrt(2*pi))~e^-frac((x-mu)^2, 2*sigma^2)), size = 5, hjust = 0) +
annotate(geom = "text", x = 1, y = 0.2, label = expression(Calculus:~italic(frac(partialdiff*N, partialdiff*t))*";"~italic(integral(f(x)*dx, a, b))), size = 5, hjust = 0) +
scale_x_continuous(limits = c(0.9, 10)) +
scale_y_continuous(limits = c(0, 1.1)) +
theme_void() +
theme(plot.background = element_rect())
We can use objects in math expressions too. However, expression() can’t handle that. Instead, we use the sister function bquote() (also in base R): the syntax is the same as that for expression(), but now we can enclose the objects in “.()” and the object values will show up in the text!
As an example, we can extract the linear model results, store them as objects, and add text annotations to the plot using bquote():
### Create functions that extract linear model results
# function for extracting coefficients
lm_coefficients_fun <- function(model){
a <- round(coef(model)[1], 2)
b <- round(coef(model)[2], 2)
if (b > 0) {
# use ".()" to substitute object values in the expression
equation <- bquote(italic(y) == .(a) + .(b)*x)
} else {
b <- abs(b)
equation <- bquote(italic(y) == .(a) - .(b)*x)
}
return(equation)
}
# function for extracting statistical results
lm_stat_results_fun <- function(model){
f <- round(summary(model)$fstatistic[1], digits = 2)
df1 <- round(summary(model)$fstatistic[2], 1)
df2 <- round(summary(model)$fstatistic[3], 1)
p_value <- round(anova(model)$`Pr(>F)`[1], 3)
r2 <- round(summary(model)$r.squared, digits = 2)
if (p_value > 0.001) {
# use ".()" to substitute object values in the expression
statistics <- bquote(italic(F[.(df1)*","*.(df2)]) == .(f)*","~italic(p) == .(p_value)*","~italic(R^2) ==.(r2))
} else {
statistics <- bquote(italic(F[.(df1)*","*.(df2)]) == .(f)*","~italic(p) < 0.001*","~italic(R^2) == .(r2))
}
return(statistics)
}
### Extract model results and store the expressions as objects
lm_iris <- lm(Sepal.Width ~ Sepal.Length, data = iris)
coefs <- lm_coefficients_fun(model = lm_iris)
stats <- lm_stat_results_fun(model = lm_iris)
### Use "bquote()" again to add the expressions to the plot
ggplot() +
annotate(geom = "text", x = 1, y = 1, label = bquote(.(coefs)), size = 5, hjust = 0) +
annotate(geom = "text", x = 1, y = 0.9, label = bquote(.(stats)), size = 5, hjust = 0) +
scale_x_continuous(limits = c(0.9, 10)) +
scale_y_continuous(limits = c(0.6, 1.1)) +
theme_void() +
theme(plot.background = element_rect())
(2) Math expressions using the package ggtext
If you have a keen eye, you might have noticed that the words in italic and bold italic in the first example look a bit funky: the letters seem to be spaced out a bit. This is totally fine, but we can make the words look “more normal” (well this depends on your perspective) using the package ggtext. The package allows users to add text annotations to ggplots using HTML syntax (check out my previous post to learn more!). Another benefit of using ggtext is that you can directly insert a line break between text, which expression() and bquote() can’t.
library(ggtext)
ggplot() +
geom_richtext(aes(x = 1, y = 1), label = "This is *italic R* and **bold R** and _**bold italic R**_", size = 5, hjust = 0, fill = NA, label.color = NA) +
geom_richtext(aes(x = 1, y = 0.8), label = "*F*<sub><span style='font-size: 8pt'>1,20</span></sub> = 5.2; *R<sup><span style='font-size: 10pt'>2</span></sup>* = 0.8", size = 5, hjust = 0, fill = NA, label.color = NA) +
geom_richtext(aes(x = 1, y = 0.6), label = "*α* = 0.05; *χ<sup><span style='font-size: 10pt'>2</span></sup>* = 3.5", size = 5, hjust = 0, fill = NA, label.color = NA) +
geom_richtext(aes(x = 1, y = 0.4), label = "This is a line break: <br> *p* = 0.001", size = 5, hjust = 0, fill = NA, label.color = NA) +
scale_x_continuous(limits = c(0.9, 10)) +
scale_y_continuous(limits = c(0, 1.1)) +
theme_void() +
theme(plot.background = element_rect())
However, ggtext has its limitations: currently it only supports a limited number of HTML tags, and therefore it’s a bit more difficult to work with math symbols (like fractions, integrals, and partial derivatives in the first example).
Summary
To recap, we learned how to add math expressions to ggplots using expression() and bquote() in base R as well as the package ggtext. Knowing the strength of each method and when to use it will give you the best outcome!
Hope you learn something useful from this post and don’t forget to leave your comments and suggestions below if you have any!
