Enable Linked Brushing via Crosstalk

Connects the chart to a Crosstalk SharedData object so that brush selections propagate to other linked widgets.

Usage

setLinked(
  myIO,
  shared_data,
  mode = "both",
  filter = FALSE,
  key = NULL,
  group = NULL,
  cursor = FALSE,
  cursorAxis = "x"
)

Arguments

myIO

an htmlwidget object created by the myIO() function

shared_data

a crosstalk::SharedData object

mode

"source", "target", or "both" (default)

filter

if TRUE, Crosstalk filter operations hide non-matching points. Default FALSE (dim only).

key

Optional character vector of row keys. When supplied, overrides the keys extracted from shared_data. Useful when the SharedData keys do not match the column used for cross-chart matching.

group

Optional character string. When supplied, overrides the Crosstalk group name from shared_data, allowing manual control over which widgets share selections.

cursor

Logical. When TRUE, a hover in any linked chart draws a synchronized crosshair on every sibling chart in the same group. Default FALSE.

cursorAxis

Character. Which axis to sync: "x" (default), "y", or "xy". Only "x" is active in v1.2.

Value

A modified myIO htmlwidget with Crosstalk linking.

Details

Selections travel on the Crosstalk key space, so a myIO chart matches rows against sibling widgets (DT, plotly, leaflet) by the same keys they use. The keys are matched to the chart's rows by position, which requires the data passed to addIoLayer() to be shared_data$data() in its original row order. If a layer's row count does not match the number of keys – for example after re-filtering the frame, or after updateMyIOData() replaced the rows – the chart falls back to matching within its own widget only, rather than pairing keys with the wrong rows.

See also

linkCharts for group-identifier linking that does not require Crosstalk (e.g. static R Markdown / Quarto HTML).

Examples

if (interactive() && requireNamespace("crosstalk", quietly = TRUE)) {
  shared <- crosstalk::SharedData$new(mtcars, key = ~rownames(mtcars))
  myIO() |>
    addIoLayer(
      type = "point", label = "scatter",
      data = shared$data(), mapping = list(x_var = "wt", y_var = "mpg")
    ) |>
    setLinked(shared)
}