Text alignment when using "expression" to make special characters in ggplot2 axis text

user_2122295 0
Viewed 327

I'm using ggplot2 to graph boxplots of patient BMIs, and I'm having trouble with the text in the x axis. I'd like to include the units for BMI (kg/m2) and I'd like the "2" to be superscripted. When I make my graph like this: 

require(plyr)
require(ggplot2)

Weights <- data.frame(SubjectID = 1:500,
                      Weight = rnorm(500, 28, 7))
Weights$Cat <- cut(Weights$Weight, breaks = c(0, 18.5, 25, 30, 40, Inf),
                   right = FALSE)
Weights$BMIcat <- revalue(Weights$Cat, 
                           c("[0,18.5)" = "Underweight\n(<18.5 kg/m2)",
                             "[18.5,25)" = "Normal weight\n(18.5 to 24.9\nkg/m2)",
                             "[25,30)" = "Overweight\n(25 to 29.9\nkg/m2)",
                             "[30,40)" = "Obese\n(30 to 39.9\nkg/m2)",
                             "[40,Inf)" = "Severely obese\n(>40 kg/m2)"))

ggplot(Weights, aes(x = BMIcat, y = Weight)) +
      geom_boxplot() +
      xlab("BMI category") + ylab("Weight (kg)")

graph with pretty axis text

everything looks BEAUTIFUL except that the 2's aren't superscripted. I'm trying to use expression to make this work, and here's the best I've come up with:

ggplot(Weights, aes(x = Cat, y = Weight)) +
      geom_boxplot() +
      scale_x_discrete(breaks=c("[0,18.5)", "[18.5,25)", "[25,30)",
                                "[30,40)", "[40,Inf)"),
                       labels=c(
                             expression("Underweight\n(<18.5 kg/m"^2*")"),
                             expression("Normal weight\n(18.5 to 24.9\nkg/m"^2*")"),
                             expression("Overweight\n(25 to 29.9\nkg/m"^2*")"),
                             expression("Obese\n(30 to 39.9\nkg/m"^2*")"),
                             expression("Severely obese\n(>40 kg/m"^2*")"))) +
      xlab("BMI category") + ylab("Weight (kg)")

ugly axis text but superscripted

Now, my 2s are superscripted but everything else looks TERRIBLE. It looks like it's suddenly left or maybe full justified instead of centered, the 2's are sometimes really far from the m's, and the axis text overlaps the graph! Any suggestions?

ggplot2 r
1 Answers

You can just use the Unicode superscript two character from the Latin-1 block

Weights$BMIcat <- revalue(Weights$Cat, 
                           c("[0,18.5)" = "Underweight\n(<18.5 kg/m²)",
                             "[18.5,25)" = "Normal weight\n(18.5 to 24.9\nkg/m²)",
                             "[25,30)" = "Overweight\n(25 to 29.9\nkg/m²)",
                             "[30,40)" = "Obese\n(30 to 39.9\nkg/m²)",
                             "[40,Inf)" = "Severely obese\n(>40 kg/m²)"))
ggplot(Weights, aes(x = BMIcat, y = Weight)) +
      geom_boxplot() +
      xlab("BMI category") + ylab("Weight (kg)")

With Unicode superscript

user_47526750
Related