tune: Hyperparameter tuning via grid or random search.

In funcml: Functional Machine Learning Framework

Hyperparameter tuning via grid or random search.

Description

Hyperparameter tuning via grid or random search.

Usage

tune(
  data,
  formula,
  model,
  grid,
  resampling = cv(5),
  metric = NULL,
  type = NULL,
  search = c("grid", "random"),
  n_evals = NULL,
  outer_resampling = NULL,
  seed = NULL,
  ncores = NULL,
  ...
)

Arguments

data	Data frame.
formula	Model formula.
model	Learner id.
grid	Data frame of hyperparameter combinations.
resampling	Resampling object.
metric	Metric to optimize.
type	Prediction type override.
search	Search strategy: "grid" or "random".
n_evals	Maximum number of configurations to evaluate when search = "random".
outer_resampling	Optional outer resampling object. When supplied, tune() performs nested resampling and reports outer-fold performance estimates for the tuned model-selection procedure.
seed	Optional seed.
ncores	Optional number of CPU cores used for tuning tasks. NULL or 1 runs sequentially.
...	Passed to fit().

Value

A funcml_tune object.

Examples

tune_obj <- tune(
  data = mtcars,
  formula = mpg ~ wt + hp,
  model = "rpart",
  grid = expand.grid(cp = c(0.001, 0.01), minsplit = c(5, 10)),
  resampling = cv(3, seed = 1),
  metric = "rmse"
)
tune_obj$best

funcml documentation built on Aug. 22, 2026, 5:08 p.m.