mlr3tuning (version 1.6.1)

tune_nested: Function for Nested Resampling

Description

Function to conduct nested resampling.

Usage

tune_nested(
  tuner,
  task,
  learner,
  inner_resampling,
  outer_resampling,
  measure = NULL,
  term_evals = NULL,
  term_time = NULL,
  terminator = NULL,
  search_space = NULL,
  store_tuning_instance = TRUE,
  store_benchmark_result = TRUE,
  store_models = FALSE,
  check_values = FALSE,
  callbacks = NULL
)

Value

mlr3::ResampleResult

Arguments

tuner

(Tuner)
Optimization algorithm.

task

(mlr3::Task)
Task to operate on.

learner

(mlr3::Learner)
Learner to tune.

inner_resampling

(mlr3::Resampling)
Resampling used for the inner loop.

outer_resampling

mlr3::Resampling)
Resampling used for the outer loop.

measure

(mlr3::Measure)
Measure to optimize. If NULL, default measure is used.

term_evals

(integer(1))
Number of allowed evaluations. Ignored if terminator is passed.

term_time

(integer(1))
Maximum allowed time in seconds. Ignored if terminator is passed.

terminator

(bbotk::Terminator)
Stop criterion of the tuning process.

search_space

(paradox::ParamSet)
Hyperparameter search space. If NULL (default), the search space is constructed from the paradox::TuneToken of the learner's parameter set (learner$param_set). When using to_tune() tokens, dependencies for hierarchical search spaces are automatically handled.

store_tuning_instance

(logical(1))
If TRUE (default), stores the internally created TuningInstanceBatchSingleCrit with all intermediate results in slot $tuning_instance. Setting store_benchmark_result = TRUE implies store_tuning_instance = TRUE, i.e. an explicit store_tuning_instance = FALSE is overridden.

store_benchmark_result

(logical(1))
If TRUE (default), store resample result of evaluated hyperparameter configurations in archive as mlr3::BenchmarkResult.

store_models

(logical(1))
If TRUE, fitted models are stored in the benchmark result (archive$benchmark_result). Setting store_models = TRUE implies store_benchmark_result = TRUE, i.e. an explicit store_benchmark_result = FALSE is overridden.

check_values

(logical(1))
If TRUE, hyperparameter values are checked before evaluation and performance scores after. If FALSE (default), values are unchecked but computational overhead is reduced.

callbacks

(list of mlr3misc::Callback)
List of callbacks.

Examples

# Nested resampling on Palmer Penguins data set
rr = tune_nested(
  tuner = tnr("random_search", batch_size = 2),
  task = tsk("penguins"),
  learner = lrn("classif.rpart", cp = to_tune(1e-04, 1e-1, logscale = TRUE)),
  inner_resampling = rsmp ("holdout"),
  outer_resampling = rsmp("cv", folds = 2),
  measure = msr("classif.ce"),
  term_evals = 2)

# Performance scores estimated on the outer resampling
rr$score()

# Unbiased performance of the final model trained on the full data set
rr$aggregate()