Skip to contents

Configure Validation for a RecursiveForecaster

Source: R/RecursiveForecaster.R
set_validate.RecursiveForecaster.Rd

Sets the $validate field of the forecaster, which controls how the validation data is constructed (see mlr3::Learner), and configures the wrapped graph so its base learner uses it (via mlr3pipelines::set_validate.GraphLearner(), the inner PipeOps receive "predefined").

Usage

# S3 method for class 'RecursiveForecaster'
set_validate(
  learner,
  validate,
  ids = NULL,
  args_all = list(),
  args = list(),
  ...
)

Arguments

learner

(RecursiveForecaster)
The forecaster to configure.

validate

(numeric(1) | "predefined" | "test" | NULL)
How to construct the internal validation data.

ids

(character() | NULL)
The ids of the PipeOps for which to enable validation, forwarded to mlr3pipelines::set_validate.GraphLearner(). Defaults to the base learner.

args_all

(named list())
Arguments passed to all set_validate() calls of the affected PipeOps.

args

(named list() of named list()s)
Arguments passed to the set_validate() calls of specific PipeOps, named by their ids.

...

(any)
Further arguments passed to mlr3pipelines::set_validate.GraphLearner().

Value

RecursiveForecaster, invisibly.