simpleml.pipelines.sklearn

Pipeline Library support for Scikit-Learn

Submodules

Package Contents

Classes

ExplicitSplitSklearnPipeline

Scikit-Learn Pipeline implementation

RandomSplitSklearnPipeline

Pipeline Wrapper with support for projected random splits on dataset

SklearnPipeline

Scikit-Learn Pipeline implementation

Attributes

__author__

simpleml.pipelines.sklearn.__author__ = Elisha Yadgaran[source]
class simpleml.pipelines.sklearn.ExplicitSplitSklearnPipeline(has_external_files=True, transformers=None, fitted=False, dataset_id=None, **kwargs)[source]

Bases: simpleml.pipelines.validation_split_mixins.ExplicitSplitMixin, simpleml.pipelines.sklearn.base.SklearnPipeline

Scikit-Learn Pipeline implementation

Parameters

  • has_external_files (bool) –

  • transformers (Optional[List[Any]]) –

  • fitted (bool) –

  • dataset_id (Optional[Union[str, uuid.uuid4]]) –

class simpleml.pipelines.sklearn.RandomSplitSklearnPipeline(train_size, test_size=None, validation_size=0.0, random_state=123, shuffle=True, **kwargs)[source]

Bases: simpleml.pipelines.validation_split_mixins.RandomSplitMixin, simpleml.pipelines.sklearn.base.SklearnPipeline

Pipeline Wrapper with support for projected random splits on dataset Useful to create a train/test/validation split on any dataset

Set splitting params: By default validation is 0.0 because it is only used for hyperparameter tuning

Parameters

  • train_size (Union[float, int]) –

  • test_size (Optional[Union[float, int]]) –

  • validation_size (Union[float, int]) –

  • random_state (int) –

  • shuffle (bool) –

class simpleml.pipelines.sklearn.SklearnPipeline(has_external_files=True, transformers=None, fitted=False, dataset_id=None, **kwargs)[source]

Bases: simpleml.pipelines.base_pipeline.Pipeline

Scikit-Learn Pipeline implementation

Parameters

  • has_external_files (bool) –

  • transformers (Optional[List[Any]]) –

  • fitted (bool) –

  • dataset_id (Optional[Union[str, uuid.uuid4]]) –

_create_external_pipeline(self, transformers, **kwargs)

Initialize a scikit-learn pipeline object

Parameters

transformers (List[Any]) –

Return type

simpleml.pipelines.sklearn.external_pipeline.SklearnExternalPipeline

_filter_fit_params(self, split)

Sklearn Pipelines register arbitrary input kwargs but validate non X,y as stepname__parameter format

Parameters

split (simpleml.pipelines.projected_splits.ProjectedDatasetSplit) –

Return type

Dict[str, Any]