R/Shiny: Group by and Summarise on column selected in dropdown? - r

Question

I am very new to R and Shiny, so I apologize in advance if this is too basic.
I am trying to render a bar graph where I select the X and Y off dropdowns. I need to aggregate by average values of Y for each X I choose. When I ran the below code, I got an error stating "Column 'y' not found". I cannot pass the actual column name selected as that can change. How do I solve this issue?
output$MultivariatePlot <- renderPlotly({
if (input$Apply > 0){
isolate({
req(data$Policies)
x <- input$MultivariateX
y <- input$MultivariateY
rv$g <- data$Policies %>%
group_by(y) %>% summarise(y = mean(y)) %>%
ggplot2::ggplot(aes_(x = sym(x), y = sym(y))) +
# ggplot2::stat_summary(fun.y = "mean", geom = "bar") +
ggplot2::geom_bar(stat='identity') +
ggplot2::scale_fill_manual(values = rara::ColorSelect(2)) +
ggplot2::theme_classic() +
ggplot2::theme(panel.grid.major = element_line(color = 'gray80', linetype = 'longdash', size = 0.3)) +
ggplot2::labs(title = paste('Comparison Between', x, 'and', y))
})

Answer 1

You can use !!sym(x) in places that you would normally use the symbol version of x, and !!sym(y) in places where you would normally use the symbol version of y.
This turns your code into the following:
rv$g <- data$Policies %>%
group_by(!!sym(x)) %>%
summarise(!!sym(y) := mean(!!sym(y))) %>%
ggplot2::ggplot(aes(x = !!sym(x), y = !!sym(y))) +
ggplot2::geom_bar(stat='identity') +
# etc
There's one last complication in there, which is that you had to use := instead of = when it's to the right of a sym(). (Also note that aes_ isn't usually recommended anymore, so I used x = !!sym(x) in regular aes(), and geom_col() is a shortcut for ggplot2::geom_bar(stat='identity')).
Here's a reproducible example of the above, which takes two strings for x and y and aggregates + plots them:
x <- "cyl"
y <- "mpg"
mtcars %>%
group_by(!!sym(x)) %>%
summarise(!!sym(y) := mean(!!sym(y))) %>%
ggplot2::ggplot(aes(x = !!sym(x), y = !!sym(y))) +
ggplot2::geom_col()