align verbosity options with documented behavior

[simonpcouch]

Closes #677. That issue points out that the verbose option is, in most cases, the option actually controlling Bayesian tuning output. This PR follows up on #547 and aims to disambiguate verbose and verbose_iter to better align with the documented functionality.

Many of the changes here transition if (control$verbose) to if (isTRUE(control$verbose_iter)), since verbose_iter isn't in control_grid() output. This also adds an iter = FALSE argument to tune_log(), a helper that is used in many different places for many different things—iter indicates that this is an update on the iterative search, which need not necessarily be an issue that triggers the catalog or log_problems().

With this PR:

library(parsnip)
library(rsample)
library(tune)
set.seed(1)

spec <- nearest_neighbor("regression", neighbors = tune())
folds <- vfold_cv(mtcars, v = 3)
init <- tune_grid(spec, mpg ~ ., folds)

res_none <- 
  tune_bayes(
    spec, mpg ~ ., folds, iter = 2, initial = init,
    control = control_bayes()
  )

res_verbose_iter <- 
  tune_bayes(
    spec, mpg ~ ., folds, iter = 2, initial = init,
    control = control_bayes(verbose_iter = TRUE)
  )
#> Optimizing rmse using the expected improvement
#> 
#> ── Iteration 1 ─────────────────────────────────────────────────────────────────
#> 
#> i Current best:      rmse=2.89 (@iter 0)
#> i Gaussian process model
#> i Generating 7 candidates
#> i Predicted candidates
#> i neighbors=1
#> ♥ Newest results:    rmse=2.85 (+/-0.289)
#> 
#> ── Iteration 2 ─────────────────────────────────────────────────────────────────
#> 
#> i Current best:      rmse=2.85 (@iter 1)
#> i Gaussian process model
#> i Generating 6 candidates
#> i Predicted candidates
#> i neighbors=3
#> ⓧ Newest results:    rmse=2.948 (+/-0.345)

res_verbose <- 
  tune_bayes(
    spec, mpg ~ ., folds, iter = 2, initial = init,
    control = control_bayes(verbose = TRUE)
  )
#> i Gaussian process model
#> ✓ Gaussian process model
#> i Generating 7 candidates
#> i Predicted candidates
#> i Estimating performance
#> i Fold1: preprocessor 1/1
#> ✓ Fold1: preprocessor 1/1
#> i Fold1: preprocessor 1/1, model 1/1
#> ✓ Fold1: preprocessor 1/1, model 1/1
#> i Fold1: preprocessor 1/1, model 1/1 (extracts)
#> i Fold1: preprocessor 1/1, model 1/1 (predictions)
#> i Fold2: preprocessor 1/1
#> ✓ Fold2: preprocessor 1/1
#> i Fold2: preprocessor 1/1, model 1/1
#> ✓ Fold2: preprocessor 1/1, model 1/1
#> i Fold2: preprocessor 1/1, model 1/1 (extracts)
#> i Fold2: preprocessor 1/1, model 1/1 (predictions)
#> i Fold3: preprocessor 1/1
#> ✓ Fold3: preprocessor 1/1
#> i Fold3: preprocessor 1/1, model 1/1
#> ✓ Fold3: preprocessor 1/1, model 1/1
#> i Fold3: preprocessor 1/1, model 1/1 (extracts)
#> i Fold3: preprocessor 1/1, model 1/1 (predictions)
#> ✓ Estimating performance
#> i Gaussian process model
#> ✓ Gaussian process model
#> i Generating 6 candidates
#> i Predicted candidates
#> i Estimating performance
#> i Fold1: preprocessor 1/1
#> ✓ Fold1: preprocessor 1/1
#> i Fold1: preprocessor 1/1, model 1/1
#> ✓ Fold1: preprocessor 1/1, model 1/1
#> i Fold1: preprocessor 1/1, model 1/1 (extracts)
#> i Fold1: preprocessor 1/1, model 1/1 (predictions)
#> i Fold2: preprocessor 1/1
#> ✓ Fold2: preprocessor 1/1
#> i Fold2: preprocessor 1/1, model 1/1
#> ✓ Fold2: preprocessor 1/1, model 1/1
#> i Fold2: preprocessor 1/1, model 1/1 (extracts)
#> i Fold2: preprocessor 1/1, model 1/1 (predictions)
#> i Fold3: preprocessor 1/1
#> ✓ Fold3: preprocessor 1/1
#> i Fold3: preprocessor 1/1, model 1/1
#> ✓ Fold3: preprocessor 1/1, model 1/1
#> i Fold3: preprocessor 1/1, model 1/1 (extracts)
#> i Fold3: preprocessor 1/1, model 1/1 (predictions)
#> ✓ Estimating performance

res_both <- 
  tune_bayes(
    spec, mpg ~ ., folds, iter = 2, initial = init,
    control = control_bayes(verbose = TRUE, verbose_iter = TRUE)
  )
#> Optimizing rmse using the expected improvement
#> 
#> ── Iteration 1 ─────────────────────────────────────────────────────────────────
#> 
#> i Current best:      rmse=2.89 (@iter 0)
#> i Gaussian process model
#> ✓ Gaussian process model
#> i Generating 7 candidates
#> i Predicted candidates
#> i neighbors=1
#> i Estimating performance
#> i Fold1: preprocessor 1/1
#> ✓ Fold1: preprocessor 1/1
#> i Fold1: preprocessor 1/1, model 1/1
#> ✓ Fold1: preprocessor 1/1, model 1/1
#> i Fold1: preprocessor 1/1, model 1/1 (extracts)
#> i Fold1: preprocessor 1/1, model 1/1 (predictions)
#> i Fold2: preprocessor 1/1
#> ✓ Fold2: preprocessor 1/1
#> i Fold2: preprocessor 1/1, model 1/1
#> ✓ Fold2: preprocessor 1/1, model 1/1
#> i Fold2: preprocessor 1/1, model 1/1 (extracts)
#> i Fold2: preprocessor 1/1, model 1/1 (predictions)
#> i Fold3: preprocessor 1/1
#> ✓ Fold3: preprocessor 1/1
#> i Fold3: preprocessor 1/1, model 1/1
#> ✓ Fold3: preprocessor 1/1, model 1/1
#> i Fold3: preprocessor 1/1, model 1/1 (extracts)
#> i Fold3: preprocessor 1/1, model 1/1 (predictions)
#> ✓ Estimating performance
#> ♥ Newest results:    rmse=2.85 (+/-0.289)
#> 
#> ── Iteration 2 ─────────────────────────────────────────────────────────────────
#> 
#> i Current best:      rmse=2.85 (@iter 1)
#> i Gaussian process model
#> ✓ Gaussian process model
#> i Generating 6 candidates
#> i Predicted candidates
#> i neighbors=3
#> i Estimating performance
#> i Fold1: preprocessor 1/1
#> ✓ Fold1: preprocessor 1/1
#> i Fold1: preprocessor 1/1, model 1/1
#> ✓ Fold1: preprocessor 1/1, model 1/1
#> i Fold1: preprocessor 1/1, model 1/1 (extracts)
#> i Fold1: preprocessor 1/1, model 1/1 (predictions)
#> i Fold2: preprocessor 1/1
#> ✓ Fold2: preprocessor 1/1
#> i Fold2: preprocessor 1/1, model 1/1
#> ✓ Fold2: preprocessor 1/1, model 1/1
#> i Fold2: preprocessor 1/1, model 1/1 (extracts)
#> i Fold2: preprocessor 1/1, model 1/1 (predictions)
#> i Fold3: preprocessor 1/1
#> ✓ Fold3: preprocessor 1/1
#> i Fold3: preprocessor 1/1, model 1/1
#> ✓ Fold3: preprocessor 1/1, model 1/1
#> i Fold3: preprocessor 1/1, model 1/1 (extracts)
#> i Fold3: preprocessor 1/1, model 1/1 (predictions)
#> ✓ Estimating performance
#> ⓧ Newest results:    rmse=2.948 (+/-0.345)

finetune is unaffected!👍

[EmilHvitfeldt]

Overall looks clean! One clarifying question

[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.