Tune hyperparameters for a model using grid search

Usage

tl_tune_grid(
  data,
  formula,
  method,
  param_grid,
  folds = 5,
  metric = NULL,
  maximize = NULL,
  verbose = TRUE,
  ...
)

Arguments

data: A data frame containing the training data
formula: A formula specifying the model
method: The modeling method to tune
param_grid: A named list of parameter values to tune
folds: Number of cross-validation folds
metric: Metric to optimize
maximize: Logical; whether to maximize (TRUE) or minimize (FALSE) the metric
verbose: Logical; whether to print progress
...: Additional arguments passed to tl_model

Value

A tidylearn model object fitted with the best hyperparameters. Tuning results are stored as an attribute "tuning_results", a list containing param_grid, results, best_params, best_metric, metric, and maximize.

results has one row per evaluated combination: mean_metric (the mean over the folds that produced a score), n_folds_ok (how many of the folds did), and a column per parameter. A parameter with a vector-valued candidate, such as hidden_layers, is a list column.

Only combinations with n_folds_ok equal to folds are eligible to be best, since a mean over the folds that happened to succeed is not comparable with a mean over all of them. If no combination completed every fold, the best of those scored on the most folds is used, with a warning. If every combination failed in every fold, the function stops.

For method = "forest", an mtry above the number of predictors is capped at that number, with a warning, and duplicate combinations that result are evaluated once.

Examples

# \donttest{
model <- tl_tune_grid(iris, Species ~ ., method = "tree",
  param_grid = list(cp = c(0.01, 0.1), minsplit = c(10, 20)),
  folds = 2, verbose = FALSE)
# }