dataeval.core.ClusterStats

class dataeval.core.ClusterStats

Pre-calculated statistics for adaptive outlier detection.

cluster_ids

Array of unique cluster IDs (excluding -1)

Type:

NDArray[np.int64]

centers

Cluster centers, shape (n_clusters, n_features)

Type:

NDArray[np.floating]

cluster_distances_mean

Mean distance from points to their cluster center, shape (n_clusters,)

Type:

NDArray[np.floating]

cluster_distances_std

Standard deviation of distances within each cluster, shape (n_clusters,)

Type:

NDArray[np.floating]

distances

Distance from each point to its nearest cluster center, shape (n_samples,)

Type:

NDArray[np.floating]

nearest_cluster_idx

Index of nearest cluster center for each point, shape (n_samples,)

Type:

NDArray[np.int64]