Integrate `latex2exp::TeX` into `add_pvalue`? about ggprism

[alesssia]

Hi @csdaw,
this is brilliant, thank you so much!

[sbaross]

Just to mention, if you replace *x* in the example expression with %*% you'll get a proper multiplication symbol × instead of a lowercase x.

stat.test$p.exprs <- paste0("P==1%*%10^", round(log10(stat.test$p.adj), 0))

[csdaw]

Hi @alesssia ,

Thanks for the detailed issue, it really helps!

As far as I know add_pvalue() can't do this, but it should be simple to add something similar which will achieve the result you're looking for without relying on latex2exp. I can give add_pvalue() a parse argument like in ggplot2::geom_text() which will be able to parse a plotmath expression.

See this simple example below using geom_text():

library(magrittr)
library(ggplot2)
library(ggprism)
# also use rstatix

p <- ggplot(ToothGrowth, aes(x = factor(dose), y = len)) + 
  geom_boxplot(aes(fill = dose), colour = "black") + 
  theme_prism() + 
  theme(legend.position = "none")

stat.test <- rstatix::t_test(ToothGrowth, len ~ dose, ref.group = "0.5") %>% 
  rstatix::add_xy_position()

## add a character column to stat.test with the plotmath expression 
## you want to use
stat.test$p.exprs <- paste0("P==1*x*10^", round(log10(stat.test$p.adj), 0))

## or use this rather ugly expression if you want everything italicised
# stat.test$p.exprs <- lapply(
#   paste(round(log10(stat.test$p.adj), 0)), 
#   function(x) bquote(italic("P = 1x10"^.(x)))
# )

p + geom_text(aes(label = p.exprs, x = xmax, y = y.position), 
              data = stat.test, parse = TRUE)

^{Created on 2021-06-15 by the reprex package (v1.0.0)}

[csdaw]

Hi @alesssia,

I've updated add_pvalue() to have a parse argument like geom_text(). You can install the Github version of ggprism to try it out.

Sorry I'm unfamiliar with latex2exp but from what I can tell it generates plotmath expression which I am a bit more familiar with. Now you can add a column to the p-value table with plotmath expressions to use to make prettier labels.

See an example below:

library(magrittr)
library(ggplot2)
library(ggprism)
# also use rstatix

p <- ggplot(ToothGrowth, aes(x = factor(dose), y = len)) + 
  geom_boxplot(aes(fill = dose), colour = "black") + 
  theme_prism() + 
  theme(legend.position = "none")

stat.test <- rstatix::t_test(ToothGrowth, len ~ dose, ref.group = "0.5") %>% 
  rstatix::add_xy_position()

stat.test$p.exprs <- paste0("P==1*x*10^", round(log10(stat.test$p.adj), 0))

p + add_pvalue(stat.test, label = "p.exprs", parse = TRUE)

^{Created on 2021-06-17 by the reprex package (v1.0.0)}