How to create a shiny reactive function with input in a separate R

Question

I am new to R and I face a problem when I trying to create a shiny application.

My app contains UI dan Server in a single app.R I want to create a package that launches a shiny app.

In the server function, I have one reactive element that subset the dataset based on dateRangeInput function from the UI. Here's part of the code in the UI

df <- coronavirus # I use coronavirus package from CRAN

ui <- sidebarLayout(
          sidebarPanel(
            dateRangeInput(
              inputId = "range",
              label = "Select date range",
              start = "2020-03-01",
              min = "2020-03-01",
              end = "2020-09-27",
              max = "2020-09-27",
              format = "dd/mm/yyyy",
              separator = " - "
            )

Originally the reactive code and output plot in Server function is

server <- function(input,output,session){
 newdate <- reactive({
     df %>% filter(between(date, input$range[1], input$range[2]))
   })

output$plotly <- renderPlotly({
    plot <- newdate() %>%
      ggplot(aes(x = date,
                 y = cases,
                 color = country)) +
      geom_line() 
ggplotly(plot)
}

I passed this reactive function into another R file named newdate.R

#' @export
newdate <- reactive({
  req(input$range)
  df %>% filter(between(date, input$range[1], input$range[2]))
})

Then I pass this newdate() function into the output plot

output$plotly <- renderPlotly({
    plot <- newdate() %>%
      ggplot(aes(x = date,
                 y = cases,
                 color = country)) +
      geom_line() 
ggplotly(plot)

There is an error that says

Error: object 'input' not found

The reactive function work fine when I put them directly in server. However when I try to put in separate R files, it gives me an error. My question is, how to put the Reactive function that has input argument into another R file? Help is much appreciated.

Waldi Waldi · Accepted Answer · 2020-10-17T06:28:50

You could pass input as argument to the reactive function:

newdate <- function(input) {
  reactive({
    req(input$range)
    df %>% filter(between(date, input$range[1], input$range[2]))
  })
}

The call to the function in the server is then:

plot <- newdate(input)() %>% ...

How to create a shiny reactive function with input in a separate R

1 Answers