`holisticai.explainability.metrics`#

Global Feature Importance

`alpha_score`	Alpha Score calculates the smallest proportion of features that account for the alpha percentage of the overall feature importance.
`fluctuation_ratio`	Calculate the fluctuation ratio for features based on partial dependencies.
`spread_divergence`	Calculates the spread divergence metric based on the inverse of the Jensen-Shannon distance (square root of the Jensen-Shannon divergence), for a given feature importance.
`rank_alignment`	Compute the rank alignment metric between conditional feature importance and ranked feature importance.
`spread_ratio`	The spread ratio, ranging from 0 to 1, measures the degree of evenness or concentration in the distribution of feature importance values.

Local Feature Importance

`importance_stability`	Determine the stability of feature importance.
`rank_consistency`	Calculate the rank consistency of local feature importances through all the instances.

Tree Based Metrics

`tree_depth_variance`	Compute the variance of the depths of the leaves in the tree (TDV).
`weighted_average_depth`	Weighted Average Depth calculates the average depth of a tree considering the number of samples that pass through each cut.
`weighted_average_explainability_score`	Weighted Average Explainability Score calculates the average depth of a tree considering the number of samples that pass through each cut.
`weighted_tree_gini`	Compute the weighted Gini index for the tree (WGNI).
`tree_number_of_rules`	Calculates the number of rules in a decision tree surrogate model.
`tree_number_of_features`	Calculates the number of features used in a decision tree surrogate model.

Surrogate Based Metrics

`surrogate_features_stability`	Calculate the stability of features used in a surrogate model.
`surrogate_fidelity_classification`	Calculate the surrogate fidelity for classification tasks.
`surrogate_fidelity_regression`	Calculate the surrogate fidelity for regression models.
`surrogate_accuracy_degradation`	Calculate the difference between the mean squared error of the original model and the surrogate model.
`surrogate_mean_squared_error_degradation`	Calculate the difference between the mean squared error of the original model and the surrogate model.