Attributes render cost to the individual marks that caused it, and splits the three phases a render goes through: build (constructing the R value), compile (the R→Rust replay, including text shaping), and raster (drawing). Turns "this plot is slow" into a specific answer.
Arguments
- scene
A
vl_scene(), or anything with anas_vellum_scene()method.- reps
Number of render repetitions to time; the median is reported.
Value
A data frame ordered by cost, with columns kind, name, n
(elements), seconds and pct (share of raster time). Rows are aggregated
per mark — a vectorised text_grob() of 200 labels compiles to 200 nodes
but reports as one row — so the table reads as "which of my marks is
expensive". The phase split is attached as the "phases" attribute and
shown when printed.
Details
Read the phase split first. Compile time is R-side — S7 property validation, grob construction, unit records — and on scenes with many small grobs it routinely dominates, in which case no amount of backend tuning will help. The per-node table only accounts for raster time.
Timing is armed only for this call, so ordinary renders pay nothing for it.
Examples
set.seed(1)
s <- vl_scene(4, 3) |>
draw(points_grob(runif(2000), runif(2000))) |>
draw(text_grob("title", y = 0.95, gp = vl_gpar(fontsize = 16)))
profile_render(s)
#> Phases (median of the timed reps):
#> • build 0.001 s (constructing the R value)
#> • compile 0.002 s (R -> Rust replay, incl. text shaping)
#> • raster 0.042 s (drawing)
#>
#> Slowest marks (raster time):
#> • circle - 2000 elem 0.0085 s 99.6%
#> • text - 1 elem 0.0000 s 0.4%