Set up auto-update/isolate logic for reactive contexts

A helper function that encapsulates the common pattern of handling auto-update functionality in module servers. When auto-update is disabled, it adds a dependency on the update button. Returns a wrapper function that either isolates reactive expressions or passes them through unchanged.

Usage

setup_auto_update_logic(input, params = NULL, track = NULL)

Arguments

input

The Shiny input object from the module server, should have both auto.update (boolean) and update (button) inputs.

params

Optional reactive-defaults store from setup_reactive_defaults(), or NULL. When supplied, an input$<key> read whose key is backed by the store resolves from the store instead of the client input, so a parameter driven by the parent app updates in a single render.

track

Optional function, or NULL. When supplied, it is called with the id of every input$<key> read through the returned wrapper, so a caller can tell which inputs an expression depends on (the modules' cached plot step keys its cache on them).

Value

A function that wraps reactive expressions. With params and track both NULL this is identity if auto-update is enabled (expressions will be reactive), or isolate if auto-update is disabled (expressions will not trigger reactivity). Otherwise it is a wrapper with the same isolation semantics that additionally redirects store-backed input$<key> reads and reports the ids read to track.

Details

This function consolidates the following common pattern:

auto_update <- input$auto.update
if (!auto_update) {
    input$update
}
isolate_fn <- if (auto_update) identity else isolate

Usage in a reactive context:

output$plot <- renderPlotly({
    isolate_fn <- setup_auto_update_logic(input)
    # Now use isolate_fn to wrap input values
    x_val <- isolate_fn(input$x.value)
})

See also

setup_reactive_defaults()

Author

Jared Andrews

Examples

if (interactive()) {
    library(shiny)
    library(plotly)

    ui <- fluidPage(
        viz_select_input("x_var", "X variable", choices = names(mtcars), selected = "wt"),
        viz_select_input("y_var", "Y variable", choices = names(mtcars), selected = "mpg"),
        checkboxInput("auto.update", "Auto-update", value = TRUE),
        actionButton("update", "Update"),
        plotlyOutput("myPlot")
    )

    server <- function(input, output, session) {
        output$myPlot <- renderPlotly({
            isolate_fn <- setup_auto_update_logic(input)
            x_val <- isolate_fn(input$x_var)
            y_val <- isolate_fn(input$y_var)
            plot_ly(mtcars, x = ~ .data[[x_val]], y = ~ .data[[y_val]], type = "scatter",
                mode = "markers")
        })
    }

    shinyApp(ui, server)
}