Recursive squarified tiling (Bruls et al.) of a value hierarchy. Leaves
carry sizes in the node table's value column (build via
as_graph() on an id/parent/value hierarchy table); parent values
are aggregated from their descendants. Every node gains
xmin/xmax/ymin/ymax within the unit square plus a leaf flag.
A derived leaves table is emitted for direct rendering with
mark_rect(data = ~leaves).
Value
A modified plotit object (pipeline form), or a new
plotit_graph when called on raw graph data.
Examples
h <- data.frame(
id = c("root", "A", "B", "a1", "a2"),
parent = c(NA, "root", "root", "A", "A"),
value = c(NA, NA, 50, 30, 20)
)
g <- as_graph(h) |> layout_treemap()
subset(g$nodes, leaf)[, c("id", "xmin", "xmax")]
#> id xmin xmax
#> 3 B 0.0 1.0
#> 4 a1 0.0 0.6
#> 5 a2 0.6 1.0
as_graph(h) |>
plotit() |>
layout_treemap() |>
mark_rect(data = ~leaves, encode(fill = id))
