Get Selected Features Per Fold

Extract the raw list of selected features for each fold from a nested CV result. Useful for debugging feature selection and examining fold-to-fold variability.

Usage

get_selected_features_per_fold(nested_result, learner_id = NULL)

Arguments

nested_result

A NestedCVResult object from BenchmarkService$run()

learner_id

Optional learner ID to filter (default: all learners)

Value

A data.table with columns: learner_id, fold, n_features, features (list column)

Examples

if (FALSE) { # \dontrun{
result <- benchmark_service$run()
features_df <- get_selected_features_per_fold(result)
print(features_df)

# View features for a specific learner/fold
features_df[learner_id == "my_learner" & fold == 1]$features[[1]]
} # }