cluster_matrix

montreal_forced_aligner.diarization.multiprocessing.cluster_matrix(ivectors, cluster_type, metric=DistanceMetric.euclidean, strict=True, no_visuals=False, working_directory=None, **kwargs)

Wrapper function for sklearn’s clustering methods

Parameters:

• ivectors (numpy.ndarray) – Ivectors to cluster

• cluster_type (ClusterType) – Clustering algorithm

• metric (DistanceMetric) – Distance metric to use in clustering

• strict (bool) – Flag for whether to raise exceptions when only one cluster is found

• kwargs – Extra keyword arguments to pass to sklearn cluster classes

Returns:

Cluster labels for each utterance

Return type:

numpy.ndarray