Skip to main content

feature_selection

feature_selection

Feature selection strategy implementations for history matching.

Classes

Name Description
AutoFeatureSelection Automatic feature selection strategy.
FeatureSelectionStrategy Abstract base class for feature selection strategies.
InteractiveFeatureSelection Interactive feature selection strategy.
ManualFeatureSelection Manual feature selection strategy.
MultiFeatureSelection Multi-feature selection strategy.

AutoFeatureSelection

feature_selection.AutoFeatureSelection(
    method='mean_sq_z',
    threshold=None,
    cooldown_period=1,
    correlation_threshold=0.8,
    max_features=1,
)

Automatic feature selection strategy.

Uses statistical metrics to automatically select the most informative features for emulation. Based on the existing feature_selection function.

Methods

Name Description
reset_history Reset the selection history (useful for testing or restarting).
select_features Select features automatically using statistical metrics.
reset_history
feature_selection.AutoFeatureSelection.reset_history()

Reset the selection history (useful for testing or restarting).

select_features
feature_selection.AutoFeatureSelection.select_features(
    simulation_results,
    observations,
    iteration=1,
)

Select features automatically using statistical metrics.

FeatureSelectionStrategy

feature_selection.FeatureSelectionStrategy()

Abstract base class for feature selection strategies.

Feature selection strategies determine which model outputs (features) should be emulated in each history matching iteration.

Methods

Name Description
get_strategy_name Return human-readable name for this strategy.
select_features Select features to emulate for this iteration.
validate_features Validate selected features exist in both simulation results and observations.
get_strategy_name
feature_selection.FeatureSelectionStrategy.get_strategy_name()

Return human-readable name for this strategy.

select_features
feature_selection.FeatureSelectionStrategy.select_features(
    simulation_results,
    observations,
    iteration=1,
)

Select features to emulate for this iteration.

Parameters
Name Type Description Default
simulation_results pd.DataFrame DataFrame with simulation outputs required
observations ObservationData ObservationData containing target observations required
iteration int Current iteration number (1-based) 1
Returns
Name Type Description
List[str] List of feature names to emulate
validate_features
feature_selection.FeatureSelectionStrategy.validate_features(
    features,
    simulation_results,
    observations,
)

Validate selected features exist in both simulation results and observations.

Parameters
Name Type Description Default
features List[str] List of feature names to validate required
simulation_results pd.DataFrame DataFrame with simulation outputs required
observations ObservationData ObservationData containing target observations required
Returns
Name Type Description
List[str] List of valid feature names
Raises
Name Type Description
ValueError If no valid features remain

InteractiveFeatureSelection

feature_selection.InteractiveFeatureSelection(fallback_strategy=None)

Interactive feature selection strategy.

Prompts the user to select features interactively. Falls back to automatic selection when running in non-interactive environments.

Methods

Name Description
select_features Select features interactively or fall back to automatic selection.
select_features
feature_selection.InteractiveFeatureSelection.select_features(
    simulation_results,
    observations,
    iteration=1,
)

Select features interactively or fall back to automatic selection.

ManualFeatureSelection

feature_selection.ManualFeatureSelection(selected_features)

Manual feature selection strategy.

Returns a predefined list of features for each iteration. Useful when you know exactly which features to emulate.

Methods

Name Description
select_features Select predefined features.
select_features
feature_selection.ManualFeatureSelection.select_features(
    simulation_results,
    observations,
    iteration=1,
)

Select predefined features.

MultiFeatureSelection

feature_selection.MultiFeatureSelection(
    n_features=2,
    method='mean_sq_z',
    correlation_threshold=0.5,
)

Multi-feature selection strategy.

Selects multiple features per iteration using automatic methods. This is a convenience wrapper around AutoFeatureSelection with max_features > 1.

Methods

Name Description
select_features Select multiple features automatically.
select_features
feature_selection.MultiFeatureSelection.select_features(
    simulation_results,
    observations,
    iteration=1,
)

Select multiple features automatically.