Using ggplot geom_text when combining superscript and variable label that contains < symbol

问题

I am having trouble adding R2 annotations to a faceted plot, where my R2 values are sometimes <0.01 (yes, it's not a good regression). I would like the 2 of R2 to be superscript. I have tried several options but seem to be stymied by the < symbol in my values

eg, using the iris data set, I first set up a new data frame with my R2 values previously calculated. The x & y positions are also set up as these are different for every facet (not essential for the iris dataset, but it is for mine)

SEr2s <- data.frame(Species = c("virginica",  "setosa",  "versicolor" ), 
                  xpos = c(5,7,7), ypos = c(4.2, 2, 4.2), 
                  lab = c("<0.01","0.08", "0.05"))

Then I run my plot:

XYPlot<-ggplot(data = iris, aes(x=Sepal.Length)) + 
  geom_point(aes(y = Sepal.Width), colour="grey40")+
  geom_smooth(aes(y = Sepal.Width), col="grey40", lwd=0.5, method="lm", se=FALSE) +
  geom_text(data = SEr2s, size=3, vjust=0, hjust=0,
            aes(x = xpos,  y = ypos, 
             label = paste("r^2==",lab)), parse = TRUE)+
  theme_bw() +
  theme(strip.background = element_blank(), strip.placement = "outside",
        panel.grid.minor = element_blank(), legend.position = "right") +
  facet_wrap(~Species)

I get this error:

Error in parse(text = text[[i]]) : :1:7: unexpected '<' 1: r^2== < ^

Is there a way to change my code or my labelling dataframe so that it doesn't try to evaluate these symbols?

回答1:

I think that the easier solution is to define the subscript in your data.frame (SEr2s):

SEr2s <- data.frame(Species = c("virginica",  "setosa",  "versicolor" ), 
                    xpos = c(5,7,7), ypos = c(4.2, 2, 4.2), 
                    lab = c("atop(R^2<0.01)","atop(R^2==0.08)", "atop(R^2==0.05)"))

Then, you can call ggplot without just with label=lab:


ggplot(data = iris, aes(x=Sepal.Length)) + 
    geom_point(aes(y = Sepal.Width), colour="grey40")+
    geom_smooth(aes(y = Sepal.Width), col="grey40", lwd=0.5, method="lm", se=FALSE) +
    geom_text(data = SEr2s, size=3, vjust=0, hjust=0,
              aes(x = xpos,  y = ypos, 
                  label = lab), parse = TRUE)+
    theme_bw() +
    theme(strip.background = element_blank(), strip.placement = "outside",
          panel.grid.minor = element_blank(), legend.position = "right") +
    facet_wrap(~Species)

I think this gives you the plot you want: plot_R2 https://ibb.co/vwbvqp2

回答2:

You can avoid plotmath if you're using the ggtext package instead, which can handle basic HTML/markdown as input.

library(ggplot2)
library(ggtext)

SEr2s <- data.frame(Species = c("virginica",  "setosa",  "versicolor" ), 
                    xpos = c(5,7,7), ypos = c(4.2, 2, 4.2), 
                    lab = c("<0.01","0.08", "0.05"))

ggplot(data = iris, aes(x=Sepal.Length)) + 
  geom_point(aes(y = Sepal.Width), colour="grey40")+
  geom_smooth(aes(y = Sepal.Width), col="grey40", lwd=0.5, method="lm", se=FALSE) +
  geom_richtext(
    data = SEr2s, size=3, vjust=0, hjust=0,
    aes(x = xpos,  y = ypos, label = paste0("r<sup>2</sup> = ", lab))
  ) +
  theme_bw() +
  theme(
    strip.background = element_blank(), strip.placement = "outside",
    panel.grid.minor = element_blank(), legend.position = "right") +
  facet_wrap(~Species)
#> `geom_smooth()` using formula 'y ~ x'

^{Created on 2019-12-02 by the reprex package (v0.3.0)}

来源：https://stackoverflow.com/questions/59129952/using-ggplot-geom-text-when-combining-superscript-and-variable-label-that-contai

标签

ggplot2

annotations

expression

ggtext