R: How to use mathematical notation in ggplot labels within a shiny app?

Question

I am trying to use mathematical notation/symbols in the labels of a ggplot2 figure within a shiny app.

The minimal example below uses shiny's built-in support for MathJax to include mathematical notation with TeX-style syntax in the user interface (UI) block of the app, as demonstrated by the \alpha symbol at the top left. However, I'd like to replace the y-axis labels in the figure with similar symbols. For example, I want to replace "a1" with the symbol produced with \alpha_1. While one can typically use latex2expr or manually create expressions for a ggplot figure, I'm not sure how to approach this when the figure is produced within shiny.

library(shiny)
library(ggplot2)
library(ggridges)

ui <- fluidPage(
  withMathJax(),
  titlePanel("Example"),
  
  sidebarLayout(
    sidebarPanel(
      numericInput(
        "alpha1",
        label = withMathJax("\\(\\alpha_{1}\\)"),
        value = 1,  min = 0, max = 500
      )),
    
    mainPanel(plotOutput("fig"))
    
  )
)

server <- function(input, output) {
  output$fig <- renderPlot({
    df <- data.frame(z = rnorm(300, mean = input$alpha1),
                     m = as.character(rep(c("a1","a2","a3"), 100)))
    ggplot(df, aes(x = z, y = m)) + geom_density_ridges()
  })
}

shinyApp(ui, server)

I tried adding scale_y_discrete to the ggplot chain, but it only gets the beta symbol correct and does not appropriately subscript.

 scale_y_discrete(labels = c(a1 = expression(paste(beta,"_{2}")), 
                             a2 = expression(withMathJax("\\(\\alpha_{1}\\)")), 
                             a3 = expression(beta)))

score 3 · Accepted Answer · answered Jul 06 '22 at 00:29

"While one can typically use latex2expr or manually create expressions for a ggplot figure, I'm not sure how to approach this when the figure is produced within shiny." It works exactly the same way.

library(shiny)
library(ggplot2)

ui <- fluidPage(

    sidebarLayout(
        sidebarPanel(
            sliderInput("bins",
                        "Number of bins:",
                        min = 1,
                        max = 50,
                        value = 30)),
        mainPanel(plotOutput("distPlot"))
    )
)

server <- function(input, output) {

    output$distPlot <- renderPlot({
        x    <- faithful[, 2]
        ggplot(data.frame(x = x), aes(x)) + 
            geom_histogram(bins = input$bins) +
            labs(
                x = expression(beta[2]),
                y = expression(alpha^2))
    })
}

shinyApp(ui = ui, server = server)

I was just answering the same thing. If you prefer to use `latex2exp`, it works too. — Fla28, Jul 06 '22 at 00:33
@Fla28 Confirmed & big fan of [`latex2exp`](https://stackoverflow.com/a/52113402/6530970) :-) — Maurits Evers, Jul 06 '22 at 00:37

R: How to use mathematical notation in ggplot labels within a shiny app?

1 Answers1