library.phases.phases_implementation.feature_analysis.feature_selection.automatic module

class library.phases.phases_implementation.feature_analysis.feature_selection.automatic.AutomaticFeatureSelection(dataset: Dataset)[source]

Bases: object

fit(type: str, max_iter: int, delete_features: bool, save_plots: bool = False, save_path: str = '')[source]
speak(message: str)[source]
class library.phases.phases_implementation.feature_analysis.feature_selection.automatic.AutomaticFeatureSelectionFactory(dataset: Dataset)[source]

Bases: ABC

abstractmethod fit(max_iter: int, delete_features: bool, save_plots: bool = False, save_path: str = '')[source]
class library.phases.phases_implementation.feature_analysis.feature_selection.automatic.BorutaAutomaticFeatureSelection(dataset: Dataset)[source]

Bases: AutomaticFeatureSelectionFactory

fit(max_iter: int = 100, delete_features: bool = True, save_plots: bool = False, save_path: str = '')[source]

Automatically selects the features that are most predictive of the target variable using the Boruta method

Parameters:

boruta_model (BorutaPy) – The Boruta model

Returns:

  • tuple

  • The predictive power features and the excluded features

class library.phases.phases_implementation.feature_analysis.feature_selection.automatic.L1AutomaticFeatureSelection(dataset: Dataset)[source]

Bases: AutomaticFeatureSelectionFactory

fit(max_iter: int = 1000, delete_features: bool = True, save_plots: bool = False, save_path: str = '')[source]

Automatically selects the features that are most predictive of the target variable using the L1 regularization method

Parameters:

  • isRegression (bool) – Whether the model is a regression model

  • print_results (bool) – Whether to print the results

Returns:

  • tuple

  • The predictive power features and the excluded features