Feature Binding Helper

GitHub Link to Code.

Feature binding helper for analysis method binding.

This module provides utilities for binding analysis methods from calculators to feature data objects with automatic data selection based on reduction state.

class mdxplain.feature.helper.feature_binding_helper.FeatureBindingHelper

Helper class for binding analysis methods to feature data.

Provides static methods for creating bound methods that automatically use the appropriate data (original or reduced) based on feature state.

static bind_stats_methods(feature_data: FeatureData) → None

Bind analysis methods from calculator to feature_data.analysis.

Creates bound methods that automatically pass the current data (reduced if available, otherwise original) to calculator analysis methods.

Parameters

feature_dataFeatureData: Feature data object to bind methods to

Returns

None: Creates feature_data.analysis object with bound methods

Examples

>>> FeatureBindingHelper.bind_stats_methods(feature_data)
>>> # Now feature_data.analysis.mean() works automatically
>>> mean_values = feature_data.analysis.mean()

static release_bound_methods(feature_data: FeatureData) → None

Release bound method references to break feature<->analysis cycles.

Parameters

feature_dataFeatureData: Feature data object whose bound analysis methods should be released.

Returns

None: Clears bound references and detaches the analysis container.

static repair_bound_methods(feature_data: FeatureData) → None

Repair BoundMethod objects after unpickling.

This method restores the feature_data and original_method references that were removed during pickling.

Parameters

feature_dataFeatureData: Feature data object with analysis attribute to repair

Returns

None: Repairs BoundMethod objects in-place