Name

adams.flow.transformer.WekaCrossValidationSplit

Synopsis

Generates train/test pairs like during a cross-validation run. It is possible to generate pairs for leave-one-out cross-validation (LOOCV) as well.
It is essential that a class attribute is set. In case of a nominal class attribute, the data gets stratified automatically.
Each of the pairs gets forwarded as a container. The training set can be accessed in the container with 'Train' and the test set with 'Test'.

Additional information

Flow input/output:
- input: weka.core.Instances
- output: adams.flow.container.WekaTrainTestSetContainer

Container information:
- adams.flow.container.WekaTrainTestSetContainer:
   - Train: training set; weka.core.Instances
   - Test: test set; weka.core.Instances
   - Seed: seed value; java.lang.Long
   - FoldNumber: current fold (1-based); java.lang.Integer
   - FoldCount: total number of folds; java.lang.Integer
   - Train original indices: original indices (0-based, train); array of int
   - Test original indices: original indices (0-based, test); array of int

Options

• loggingLevel
  

  The logging level for outputting errors and debugging output.

  command-line	-logging-level <OFF|SEVERE|WARNING|INFO|CONFIG|FINE|FINER|FINEST>
  default	WARNING
  min-user-mode	Expert

  
  
• name
  

  The name of the actor.

  command-line	-name <java.lang.String>
  default	WekaCrossValidationSplit

  
  
• annotations
  

  The annotations to attach to this actor.

  command-line	-annotation <adams.core.base.BaseAnnotation>
  default

  
  
• skip
  

  If set to true, transformation is skipped and the input token is just forwarded as it is.

  command-line	-skip <boolean>
  default	false

  
  
• stopFlowOnError
  

  If set to true, the flow execution at this level gets stopped in case this actor encounters an error; the error gets propagated; useful for critical actors.

  command-line	-stop-flow-on-error <boolean>
  default	false
  min-user-mode	Expert

  
  
• silent
  

  If enabled, then no errors are output in the console; Note: the enclosing actor handler must have this enabled as well.

  command-line	-silent <boolean>
  default	false
  min-user-mode	Expert

  
  
• seed
  

  The seed value for the randomization; overrides the value defined by the fold generator scheme.

  command-line	-seed <long>
  default	1

  
  
• folds
  

  The number of folds to use in the cross-validation; use -1 for leave-one-out cross-validation (LOOCV); overrides the value defined by the fold generator scheme.

  command-line	-folds <int>
  default	10

  
  
• relationName
  

  The placeholders for creating the relation name: @ = original relation name, $T = type (train/test), $N = current fold number; overrides the value defined by the fold generator scheme.

  command-line	-relation <java.lang.String>
  default	@

  
  
• createView
  

  If enabled, views of the dataset are created instead of actual copies; overrides the value defined by the fold generator scheme.

  command-line	-create-view <boolean>
  default	false

  
  
• generator
  

  The scheme to use for generating the folds; the actor options take precedence over the scheme's ones.

  command-line	-generator <weka.classifiers.CrossValidationFoldGenerator>
  default	weka.classifiers.DefaultCrossValidationFoldGenerator