Extract the candidates a final fit scored

extract_scored_candidates() returns the parameter settings that the tuning run of nested_final_fit() scored, one row per candidate. Compare them with collect_inner_metrics() on the nested run.

Usage

extract_scored_candidates(x, ...)

Arguments

x

A nested_final_fit from nested_final_fit().

...

Not used, and must be empty.

Value

A tibble with one column per tuned parameter, then .config, and .iter where the search iterated. A candidate scored at several evaluation times has one row. A final fit that ran no tuning gives an error.

Details

tune can reach fewer candidates than a grid size asks for. A candidate that failed on every resample scored nothing and is missing here. tune::collect_notes(extract_tune_results(x)) shows why it failed.

See also

extract_tune_results(), nested_final_fit(), nested_tune_grid()

Examples

data(mtcars)

rec <- recipes::recipe(mpg ~ ., data = mtcars) |>
  recipes::step_pca(recipes::all_predictors(), num_comp = tune::tune())
wf <- workflows::workflow(rec, parsnip::linear_reg())

set.seed(1)
folds <- nested_resamples(
  mtcars,
  outside = rsample::vfold_cv(v = 2),
  inside = rsample::vfold_cv(v = 2)
)
set.seed(2)
res <- nested_tune_grid(wf, folds, grid = data.frame(num_comp = 1:2))
set.seed(3)
final <- nested_final_fit(wf, res)
extract_scored_candidates(final)
#> # A tibble: 2 × 2
#>   num_comp .config        
#>      <int> <chr>          
#> 1        1 pre1_mod0_post0
#> 2        2 pre2_mod0_post0