Feature Reset Helper

GitHub Link to Code.

Feature reset helper for managing feature data cleanup.

This module provides utilities for resetting and clearing feature data in the FeatureManager, handling both single and multiple feature resets with different error handling modes.

class mdxplain.feature.helper.feature_reset_helper.FeatureResetHelper

Helper class for feature reset operations.

Provides static methods for handling feature reset logic, including selective resets, error handling modes, and status reporting.

static reset_all_features(pipeline_data: PipelineData) → None

Reset all features in the pipeline data.

Parameters

pipeline_dataPipelineData: Pipeline data object to reset

Returns

None: Clears all feature data and prints summary

Examples

>>> FeatureResetHelper.reset_all_features(pipeline_data)
Reset 3 feature(s): distances, contacts, angles
All feature data has been cleared. Features must be recalculated.

static reset_specific_features(pipeline_data: PipelineData, feature_types: List[Any], strict: bool = False) → None

Reset specific feature types with optional strict error handling.

Parameters

pipeline_dataPipelineData: Pipeline data object to reset
feature_typesList[Any]: List of feature types to reset (strings, instances, or classes)
strictbool, default=False: Whether to raise errors for non-existent features

Returns

None: Resets specified features and prints status

Raises

ValueError: If strict=True and feature types are not found

Examples

>>> FeatureResetHelper.reset_specific_features(
...     pipeline_data, ["distances", "contacts"], get_key_fn
... )
Reset 2 feature(s): distances, contacts

static normalize_feature_types(feature_type: Any | List[Any]) → List[Any]

Normalize feature_type parameter to list format.

Parameters

feature_typeFeatureType or List[FeatureType]: Single feature type or list of feature types

Returns

List[Str]: Normalized list of feature types

Examples

>>> types = FeatureResetHelper.normalize_feature_types("distances")
>>> print(types)
["distances"]

>>> types = FeatureResetHelper.normalize_feature_types(["distances", "contacts"])
>>> print(types)
["distances", "contacts"]

static check_has_features(pipeline_data: PipelineData) → bool

Check if pipeline data has any features to reset.

Parameters

pipeline_dataPipelineData: Pipeline data object to check

Returns

bool: True if features exist, False otherwise

Examples

>>> has_features = FeatureResetHelper.check_has_features(pipeline_data)
>>> if not has_features:
...     print("No features to reset.")