`myIOProxy()` creates a lightweight handle to an already-rendered myIO widget, and `updateMyIOData()` swaps the data of one or more existing layers without re-rendering the whole widget. Unlike re-executing `renderMyIO()` (which destroys and recreates the chart on every reactive invalidation, dropping brush/zoom/toggle state and flickering), a proxy update reuses the existing data-join path: only the changed marks transition, and interaction state is preserved.
Arguments
- outputId
The output id of the `myIOOutput()` whose chart to update.
- session
The Shiny session object. Defaults to the current reactive domain.
- proxy
A `myIO_proxy` object from `myIOProxy()`.
- ...
One or more `label = data.frame` updates, where `label` is an existing layer label and the data frame carries that layer's mapped columns.
Details
Layers are matched by their `label`. Unknown labels are ignored client-side. The supplied data frame replaces the layer's data as-is (the identity data path); statistical transforms attached at `addIoLayer()` time are not re-applied, so pass already-transformed data for transformed layers.
Examples
if (FALSE) { # \dontrun{
library(shiny)
ui <- fluidPage(myIOOutput("chart"), actionButton("go", "Resample"))
server <- function(input, output, session) {
output$chart <- renderMyIO({
myIO(data = data.frame(x = 1:50, y = rnorm(50))) |>
addIoLayer("line", label = "series", mapping = list(x_var = "x", y_var = "y"))
})
observeEvent(input$go, {
myIOProxy("chart") |>
updateMyIOData(series = data.frame(x = 1:50, y = rnorm(50)))
})
}
shinyApp(ui, server)
} # }