holisticai.explainability.metrics.weighted_tree_gini#

holisticai.explainability.metrics.weighted_tree_gini(tree)[source]#

Compute the weighted Gini index for the tree (WGNI). Reference value: 0.0

Parameters

treeTree: The tree to compute the weighted Gini index of.

Returns

float: The weighted Gini index of the tree.

Examples

>>> from sklearn.datasets import load_iris
>>> from sklearn.tree import DecisionTreeClassifier
>>> from holisticai.explainability.metrics import weighted_average_depth
>>> X, y = load_iris(return_X_y=True)
>>> clf = DecisionTreeClassifier()
>>> clf.fit(X, y)
>>> weighted_tree_gini(clf.tree_)