shiny debounce doesn't render initial plot - r

Question

If I add a debounce to a get_data() reactive expression, the first time the data is retrieved the plot does not render. However, changing the data (by selecting a new mpg) causes the plot to then render. Why is this? Is there a workaround?
Here is a simple minimal example demonstrating the problem. Try removing %>% debounce(500) to see that it works as expected without it:
if (interactive()) {
options(device.ask.default = FALSE)
library(shiny)
library(magrittr)
ui <- fluidPage(
selectInput("select", label = "select mpg", choices = c(mtcars$mpg, ""), selected = ""),
plotOutput("plot")
)
server <- function(input, output, session) {
get_data <- reactive({
req(input$select)
mtcars[mtcars$mpg == input$select,]
}) %>% debounce(500)
get_plot <- reactive({
data <- get_data()
print(data)
plot(get_data())
})
output$plot <- renderPlot({
get_plot()
})
}
shinyApp(ui, server)
}

Answer 1

Couple of things going on here. I don't think we are allowed to have duplicates in a select input. mtcars$mpg has duplicate values inside it. Setting the initial value to "" is also causing strange behaviors. If you really want the initial plot to be empty along with debounce we could set it to " ". Here is what that would look like.
if (interactive()) {
options(device.ask.default = FALSE)
library(shiny)
library(magrittr)
ui <- fluidPage(
selectInput("select", label = "select mpg", choices = c(" ",unique(mtcars$mpg)),selected = " "),
plotOutput("plot")
)
server <- function(input, output, session) {
get_data <- reactive({
req(input$select)
if(!is.na(as.numeric(input$select))) mtcars[mtcars$mpg == input$select,] else NULL
}) %>% debounce(500)
get_plot <- reactive({
req(get_data())
data <- get_data()
print(data)
plot(get_data())
})
output$plot <- renderPlot({
get_plot()
})
}
shinyApp(ui, server)
}
Else if you're okay with having an initial plot the following works. Note its important to use unique()
if (interactive()) {
options(device.ask.default = FALSE)
library(shiny)
library(magrittr)
ui <- fluidPage(
selectInput("select", label = "select mpg", unique(mtcars$mpg)),
plotOutput("plot")
)
server <- function(input, output, session) {
get_data <- reactive({
req(input$select)
mtcars[mtcars$mpg == input$select,]
}) %>% debounce(500)
get_plot <- reactive({
req(get_data())
data <- get_data()
print(data)
plot(get_data())
})
output$plot <- renderPlot({
get_plot()
})
}
shinyApp(ui, server)
}
I even tried replacing the select input with selectizeInput("select", label = "select mpg", choices = unique(mtcars$mpg),multiple = TRUE,options = list(maxItems = 1)) This still caused issues.