interdim.score

Functions

score_clustering(X, labels[, true_labels, ...])

Evaluate clustering performance using the specified method.
interdim.score.score_clustering(X: ndarray, labels: ndarray, true_labels: ndarray | None = None, method: Literal['adjusted_mutual_info', 'adjusted_rand', 'completeness', 'fowlkes_mallows', 'homogeneity', 'mutual_info', 'normalized_mutual_info', 'rand', 'v_measure', 'contingency_matrix', 'pair_confusion_matrix', 'calinski_harabasz', 'davies_bouldin', 'silhouette'] = 'silhouette') → float | ndarray[source]

Evaluate clustering performance using the specified method.

Parameters:

  • X – The input data.

  • labels – The predicted cluster labels.

  • true_labels – The true cluster labels (optional).

  • method – The evaluation method to use.

Returns:

The computed metric (float or array).

Raises:

ValueError – If an unsupported evaluation method is specified or if true labels are not provided for methods that require them.