.Last.tune.result contains intermediate results

[simonpcouch]

library(tidymodels)

svm_mod <- svm_rbf(mode = "regression", cost = tune()) %>%
  set_engine("kernlab")

set.seed(7898)
data_folds <- vfold_cv(mtcars, v = 2)

rec <- recipe(mpg ~ ., data = mtcars) %>%
  step_bs(disp, deg_free = 1)

cars_res <- tune_bayes(svm_mod, preprocessor = rec,
                        resamples = data_folds)
#> → 1 | warning: some 'x' values beyond boundary knots may cause ill-conditioned bases
#> There were issues with some computations   1: x15

cars_res
#> # Tuning results
#> # 2-fold cross-validation 
#> # A tibble: 22 × 5
#>    splits          id    .metrics          .notes           .iter
#>    <list>          <chr> <list>            <list>           <int>
#>  1 <split [16/16]> Fold1 <tibble [10 × 5]> <tibble [0 × 3]>     0
#>  2 <split [16/16]> Fold2 <tibble [10 × 5]> <tibble [5 × 3]>     0
#>  3 <split [16/16]> Fold1 <tibble [2 × 5]>  <tibble [0 × 3]>     1
#>  4 <split [16/16]> Fold2 <tibble [2 × 5]>  <tibble [1 × 3]>     1
#>  5 <split [16/16]> Fold1 <tibble [2 × 5]>  <tibble [0 × 3]>     2
#>  6 <split [16/16]> Fold2 <tibble [2 × 5]>  <tibble [1 × 3]>     2
#>  7 <split [16/16]> Fold1 <tibble [2 × 5]>  <tibble [0 × 3]>     3
#>  8 <split [16/16]> Fold2 <tibble [2 × 5]>  <tibble [1 × 3]>     3
#>  9 <split [16/16]> Fold1 <tibble [2 × 5]>  <tibble [0 × 3]>     4
#> 10 <split [16/16]> Fold2 <tibble [2 × 5]>  <tibble [1 × 3]>     4
#> # … with 12 more rows
#> 
#> There were issues with some computations:
#> 
#>   - Warning(s) x15: some 'x' values beyond boundary knots may cause ill-conditioned b...
#> 
#> Run `show_notes(.Last.tune.result)` for more information.

.Last.tune.result
#> # Tuning results
#> # 2-fold cross-validation 
#> # A tibble: 2 × 4
#>   splits          id    .metrics         .notes          
#>   <list>          <chr> <list>           <list>          
#> 1 <split [16/16]> Fold1 <tibble [2 × 5]> <tibble [0 × 3]>
#> 2 <split [16/16]> Fold2 <tibble [2 × 5]> <tibble [1 × 3]>
#> 
#> There were issues with some computations:
#> 
#>   - Warning(s) x1: some 'x' values beyond boundary knots may cause ill-conditioned b...
#> 
#> Run `show_notes(.Last.tune.result)` for more information.

^{Created on 2023-02-17 with reprex v2.0.2}

[simonpcouch]

This happens because tune_bayes() iteratively calls tune_grid() as part of the GP. That function "resets" the last result each time:

tune/R/tune_grid.R

Lines 298 to 307 in c74dfe4

	res <-
	tune_grid_workflow(
	object,
	resamples = resamples,
	grid = grid,
	metrics = metrics,
	pset = param_info,
	control = control
	)
	.stash_last_result(res)

[github-actions]

This issue 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.