speed up tune_tbl() helper

[simonpcouch]

tune_tbl() is called each resample fit, and is a light wrapper around tibble::tibble(). The function is only called inside of tune_args(), so we know what kinds of inputs the helper will receive:

parsnip/R/tune_args.R

Lines 21 to 29 in d8f273c

	tune_tbl(
	name = names(res),
	tunable = unname(!is.na(res)),
	id = res,
	source = "model_spec",
	component = model_type,
	component_id = NA_character_,
	full = full
	)

bm <- 
  bench::mark(
    total = fit_resamples(linear_reg(), mpg ~ ., bootstraps(mtcars, 100)),
    tune_tbl = replicate(100, parsnip:::tune_tbl()),
    check = FALSE
  )

bm
#> # A tibble: 2 × 6
#>   expression      min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr> <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 total         3.25s    3.25s     0.308    47.7MB     8.92
#> 2 tune_tbl    43.77ms  47.91ms    22.6        77KB    11.3

As a percentage of total time, the helper takes:

100 * as.numeric(bm$median[2]) / as.numeric(bm$median[1])
#> [1] 1.474154

Since we can recycle vectors as needed before inputting to the helper, we can use new_tibble(list()) instead of tibble(), which will drastically cut down on the time this helper takes.

[EmilHvitfeldt]

Looks clean!

[github-actions]

This pull request has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.