simpleml.transformers

Base Module for Transformers

Use Mixin class for expected methods in multiple inheritance subclasses. Use Base class for other ones

Subpackages

Submodules

Package Contents

Classes

Transformer

Base Transformer class with param management - Can interfere with mro

TransformerMixin

Base Transformer class that implements all the necessary methods

Attributes

__author__

simpleml.transformers.__author__ = Elisha Yadgaran[source]
class simpleml.transformers.Transformer(**kwargs)[source]

Bases: TransformerMixin

Base Transformer class with param management - Can interfere with mro if used as a mixin - Use TransformerMixin in that case

Assumes only seeding kwargs passed - will affect hash otherwise if random unused parameters are passed

get(self, param)
Parameters

param (str) –

Return type

Any

get_params(self, **kwargs)

Should only return seeding parameters, not fit ones (ie params of unfit object should be identical to fit object)

Return type

Dict[str, Any]

set_params(self, **kwargs)
Return type

None

class simpleml.transformers.TransformerMixin[source]

Bases: sklearn.base.TransformerMixin

Base Transformer class that implements all the necessary methods

Default behavior is to do nothing - overwrite later

object_type :str = TRANSFORMER
fit(self, X, y=None, **kwargs)
Parameters

  • X (Any) –

  • y (Optional[Any]) –

get_feature_names(self, input_feature_names)
Parameters

input_feature_names (List[str]) –

Return type

List[str]

get_params(self, **kwargs)

Should only return seeding parameters, not fit ones (ie params of unfit object should be identical to fit object)

Return type

Dict[str, Any]

abstract partial_fit(self, X, y=None, **kwargs)
Parameters

  • X (Any) –

  • y (Optional[Any]) –

abstract reset(self)

Flag to reset state to initialized values. Expects subclasses to implement

Return type

None

set_params(self, **kwargs)
Return type

None

transform(self, X, y=None, **kwargs)
Parameters

  • X (Any) –

  • y (Optional[Any]) –

Return type

Any