Feature Computation Helper

GitHub Link to Code.

Feature computation helper for managing feature calculation workflow.

This module provides utilities for handling feature computation processes, including existence checks, dependency validation, and result storage in the FeatureManager.

class mdxplain.feature.helper.feature_computation_helper.FeatureComputationHelper

Helper class for feature computation operations.

Provides static methods for managing the feature computation workflow, including pre-computation checks, computation execution, and result storage.

static check_feature_existence(pipeline_data: PipelineData, feature_key: str, traj_indices: List[int], force: bool) → None

Check if feature already exists and handle accordingly.

Parameters

pipeline_dataPipelineData: Pipeline data object to check
feature_keystr: Feature key to check
forcebool: Whether to force recomputation

Returns

None: Handles existing features or raises ValueError

Raises

ValueError: If feature exists and force=False

Examples

>>> FeatureComputationHelper.check_feature_existence(
...     pipeline_data, "distances", False
... )

static execute_computation(pipeline_data: PipelineData, feature_data: FeatureData, feature_type: FeatureTypeBase, traj_idx: int, force_original: bool = True) → Tuple[Any, Dict]

Execute feature computation for a single trajectory.

Parameters

pipeline_dataPipelineData: Pipeline data object with input data
feature_dataFeatureData: Feature data object for computation
feature_typeFeatureTypeBase: Feature type object to compute
traj_idxint: Trajectory index to process
force_originalbool, default=True: Whether to force using original data instead of reduced data

Returns

Tuple[Any, Dict]: Tuple of (computed_data, metadata) for single trajectory

Examples

>>> # Process single trajectory
>>> data, metadata = FeatureComputationHelper.execute_computation(
...     pipeline_data, feature_data, distances_feature, traj_idx=0
... )

static store_computation_results(feature_data: FeatureData, data: ndarray, feature_metadata: Dict) → None

Store computation results in feature data object.

Parameters

feature_dataFeatureData: Feature data object to store results in
datanp.ndarray: Computed feature data (typically numpy array)
feature_metadataDict: Feature metadata dictionary

Returns

None: Stores data and metadata in feature_data object

Examples

>>> FeatureComputationHelper.store_computation_results(
...     feature_data, distance_matrix, metadata_dict
... )