Nested Cross-Validation

This implements the Nested CV resampling procedure by Bates et al. (2024).

Parameters

• folds :: integer(1)
  The number of folds. This is initialized to 5.

• repeats :: integer(1)
  The number of repetitions. THis is initialized to 10.

References

Bates, Stephen, Hastie, Trevor, Tibshirani, Robert (2024). “Cross-validation: what does it estimate and how well does it do it?” Journal of the American Statistical Association, 119(546), 1434–1445.

Super class

mlr3::Resampling -> ResamplingNestedCV

Active bindings

iters

(integer(1))
The total number of resampling iterations.

Methods

Public methods

• ResamplingNestedCV$new()

• ResamplingNestedCV$unflatten()

• ResamplingNestedCV$clone()

Inherited methods

• mlr3::Resampling$format()
• mlr3::Resampling$help()
• mlr3::Resampling$instantiate()
• mlr3::Resampling$print()
• mlr3::Resampling$test_set()
• mlr3::Resampling$train_set()

---

Method new()

Creates a new instance of this R6 class.

Usage

ResamplingNestedCV$new()

---

Method unflatten()

Convert a resampling iteration to a more useful representation. For outer resampling iterations, inner is NA.

Usage

ResamplingNestedCV$unflatten(iter)

Arguments

iter

(integer(1))
The iteration.

Returns

list(rep, outer, inner)

---

Method clone()

The objects of this class are cloneable with this method.

Usage

ResamplingNestedCV$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

ncv = rsmp("nested_cv", folds = 3, repeats = 10L)
ncv
#> <ResamplingNestedCV>: Nested CV
#> * Iterations: 90
#> * Instantiated: FALSE
#> * Parameters: folds=3, repeats=10
rr = resample(tsk("mtcars"), lrn("regr.featureless"), ncv)