Add interactivity for geom_pcp

It would be elegant to be able to create interactive parallel coordinate plots with plotly layered on the ggpcp library. Users should be able to select a line (or lines) and the entire line would highlight.

library(ggplot2)
data(mtcars)
mtcars_pcp <- mtcars |>
  dplyr::mutate(
    cyl = factor(cyl),
    vs = factor(vs),
    am = factor(am),
    gear = factor(gear),
    carb = factor(carb)
  ) |>
  pcp_select(1:11) |>  # select everything
  pcp_scale() |>
  pcp_arrange()

 base <- mtcars_pcp |> ggplot(aes_pcp())


 # Just the base plot:
 base + geom_pcp()

# Ideally this would allow lines to be selected 
ggplotly() 

# Or something like:
highlight_key(mtcars) |> ... (ggplot2 code) |> ggplotly()

Start with the geom_pcp and ggplotly entry points described in the issue, and compare the proposed highlight_key approach with the ggpcp pipeline using pcp_select, pcp_scale, and pcp_arrange. Done means a parallel coordinate plot supports selecting one or more lines and highlights the entire selected line interactively.