lale.lib.sklearn.target_encoder module

class lale.lib.sklearn.target_encoder.TargetEncoder(*, categories='auto', target_type='auto', smooth='auto', cv=5, shuffle=True, random_state=None)

Bases: PlannedIndividualOp

Target encoder for regression and classification targets..

This documentation is auto-generated from JSON schemas.

Parameters

• categories (union type, not for optimizer, default 'auto') –

  Categories (unique values) per feature.

  • ’auto’

    Determine categories automatically from training data.

  • or array

    The ith list element holds the categories expected in the ith column.

    • items : union type

      • array of items : string

      • or array of items : float

        Should be sorted.

• target_type (union type, not for optimizer, default 'auto') –

  Type of target.

  • ’auto’

    Type of target is inferred with type_of_target.

  • or ‘continuous’

    Continuous target

  • or ‘binary’

    Binary target

  • or ‘multiclass’

    Multiclass target

• smooth (union type, optional, not for optimizer, default 'auto') –

  The amount of mixing of the target mean conditioned on the value of the category with the global target mean.

  • ’auto’

    Set to an empirical Bayes estimate.

  • or float, >=0.0, <=1.0

    A larger smooth value will put more weight on the global target mean

• cv (integer, >=1, optional, not for optimizer, default 5) – Determines the number of folds in the cross fitting strategy used in fit_transform. For classification targets, StratifiedKFold is used and for continuous targets, KFold is used.

• shuffle (boolean, optional, not for optimizer, default True) – Whether to shuffle the data in fit_transform before splitting into folds. Note that the samples within each split will not be shuffled.

• random_state (union type, optional, not for optimizer, default None) –

  When shuffle is True, random_state affects the ordering of the indices, which controls the randomness of each fold. Otherwise, this parameter has no effect. Pass an int for reproducible output across multiple function calls.

  • None

  • or numpy.random.RandomState

    Use the provided random state, only affecting other users of that same random state instance.

  • or integer

    Explicit seed.

fit(X, y=None, **fit_params)

Train the operator.

Note: The fit method is not available until this operator is trainable.

Once this method is available, it will have the following signature:

Parameters

• X (array) –

  Features; the outer array is over samples.

  • items : union type

    • array of items : float

    • or array of items : string

• y (array, optional) – The target data used to encode the categories.

transform(X, y=None)

Transform the data.

Note: The transform method is not available until this operator is trained.

Once this method is available, it will have the following signature:

Parameters

X (array) –

Features; the outer array is over samples.

• items : union type

  • array of items : float

  • or array of items : string

Returns

result – Transformed input; the outer array is over samples.

• items : union type

  • array of items : float

  • or array of items : string

Return type

array