R Shiny: Can't access reactive value outside of reactive consumer

Question

I am trying to build a Shiny app that performs some modifications to a raster file, plots it and gives option to download (modified) raster files. I am getting the following error:

Listening on http://127.0.0.1:3371
Warning: Error in : Can't access reactive value 'divider' outside of reactive consumer.
ℹ Do you need to wrap inside reactive() or observer()?
  53: <Anonymous>
Error : Can't access reactive value 'divider' outside of reactive consumer.
ℹ Do you need to wrap inside reactive() or observer()?

What am I doing wrong? I found this related question on RStudio community but it didn't help me. Could anyone please help me understand this?

Minimum Working Example

ui.R

shinyUI(fluidPage(

    # Application title
    titlePanel("Urban-Rural Raster Population"),

    # Sidebar 
    sidebarLayout(
        sidebarPanel(
            sliderInput("divider",
                        "Divider number:",
                        min = 1,
                        max = 29,
                        value = 7),
            downloadButton("download1", "Download Rural Raster"),
            downloadButton("download2", "Download Urban Raster")
        ),

        # plots
        mainPanel(
            plotOutput("plot1"),
            plotOutput("plot2")
        )
    )
))

server.R

# Loading all external data - so that its done only once
rast_pop = read_stars("ind_ppp_2020_1km_Aggregated_UNadj.tif", proxy = F)
urca_raster = read_stars("URCA.tif", proxy = T)
names(urca_raster) = "urca"

# Finding rasters for India's boundaries
ind_shp = st_as_sf(getData('GADM', country ='IND', level = 0)) # for country
ind_st_shp = st_as_sf(getData('GADM', country ='IND', level = 1))  #for states boundary
ind_dis_shp = st_as_sf(getData('GADM', country = 'IND', level = 2)) # for district boundary

# cropping India from raster
ind_urca = st_crop(urca_raster, ind_shp)


# Define server logic 
shinyServer(function(input, output) {

    div_ru = input$divider
    
    # dividing on the basis of rural and urban
    ind_rur = ind_urca > div_ru
    ind_urb = ind_urca <= div_ru
    ind_rur_star = st_as_stars(ind_rur)
    ind_urb_star = st_as_stars(ind_urb)
    
    output$plot1 = renderPlot({
        plot(ind_rur_star)
        })
    
    output$plot2 = renderPlot({
        plot(ind_urb_star)
    })
    
    output$download1 = downloadHandler(
        filename = function() {
            paste0("rural_raster_", input$divider, ".tif")
            },
        content = function(file) {
            write_stars(ind_rur_star, file, layer = 1)
        }
    )
    output$download2 = downloadHandler(
        filename = function() {
            paste0("urban_raster_", input$divider, ".tif")
        },
        content = function(file) {
            write_stars(ind_urb_star, file, layer = 1)
        }
    )
})

Answer 1

The problem is that, as the error suggests, divider is reactive, but you are using it outside of the reactive environment. For example, variables like ind_rur need to be recomputed each time divider changes, but in your code they are computed only once.

Try this for your server function:

shinyServer(function(input, output) {

    div_ru <- reactive({
       input$divider
    })
    
    # dividing on the basis of rural and urban
    ind_rur <- reactive({ind_urca > div_ru()})
    ind_urb <- reactive({ind_urca <= div_ru()})
    ind_rur_star <- reactive({st_as_stars(ind_rur())})
    ind_urb_star <- reactive({st_as_stars(ind_urb())})
    
    output$plot1 = renderPlot({
        plot(ind_rur_star())
        })
    
    output$plot2 = renderPlot({
        plot(ind_urb_star())
    })
    
    output$download1 = downloadHandler(
        filename = function() {
            paste0("rural_raster_", div_ru(), ".tif")
            },
        content = function(file) {
            write_stars(ind_rur_star(), file, layer = 1)
        }
    )
    output$download2 = downloadHandler(
        filename = function() {
            paste0("urban_raster_", div_ru(), ".tif")
        },
        content = function(file) {
            write_stars(ind_urb_star(), file, layer = 1)
        }
    )
})