Skip to contents

Attach annotation strips to a base plot

Source: R/compose.R
compose_annot.Rd

compose_annot() glues one to four plotit strips (dendrograms, annotation bars, marginal densities) onto the matching sides of a base plot. Strips share the base panel span: patchwork aligns the panel areas of plots stacked in the same design row/column, so tree tips and annotation cells line up with the base heatmap rows/columns by construction.

Usage

compose_annot(
  base,
  top = NULL,
  bottom = NULL,
  left = NULL,
  right = NULL,
  heights = NULL,
  widths = NULL,
  gap = 0,
  guides = "collect",
  align = "panel",
  on_top = FALSE
)

Arguments

base

A plotit or plotit_composite object.

top, bottom, left, right

Optional plotit/plotit_composite strips attached to the matching side.

heights, widths

Strip sizes for the vertical (top/bottom) and horizontal (left/right) sides: NULL (default) = 0.9 inch per strip; grid::unit() = absolute size; numeric = ratio relative to the base row/column (same three-state contract as compose_grid()).

gap

Space between the strips and the base plot in inches (default 0; realised as empty spacer rows/columns).

guides

"collect" (default) to merge legends, "keep" to keep per-panel legends.

align

Alignment reference. "panel" (default) = strips align to the base panel area. "plot" requests full-area alignment; grid members always align on panel areas, so it currently falls back to "panel" (documented approximation).

on_top

Reserved for overlap semantics (use compose_inset()); TRUE warns and renders the strips beside the base.

Value

A plotit_composite object.

Details

This is the same assembly engine behind compose_marginal() (marginal = a top + right special case with axis-hiding conventions). The result can be nested into compose_grid().

Examples

mat <- matrix(c(9, 1, 8, 2, 1, 9, 2, 8, 5, 3, 7, 4),
  nrow = 4,
  dimnames = list(paste0("g", 1:4), paste0("s", 1:3))
)
h <- stats::hclust(stats::dist(mat))
hm <- plotit(mat, encode()) |> mark_heatmap(cluster = h)
tree <- as_graph(h) |>
  plotit() |>
  layout_dendrogram(direction = "up") |>
  mark_rule(data = ~edges)
hm |> compose_annot(top = tree)