scib_metrics.silhouette_batch

scib_metrics.silhouette_batch(X, labels, batch, rescale=True, chunk_size=256, metric='euclidean', between_cluster_distances='nearest')

Average silhouette width (ASW) with respect to batch ids within each label [Luecken et al., 2022].

Default parameters (‘euclidean’, ‘nearest’) match scIB implementation.

Additional options enable BRAS compatible usage (see bras() documentation).

Parameters:

X ndarray

Array of shape (n_cells, n_features).

labels ndarray

Array of shape (n_cells,) representing label values

batch ndarray

Array of shape (n_cells,) representing batch values

rescale bool (default: True)

Scale asw into the range [0, 1]. If True, higher values are better.

chunk_size int (default: 256)

Size of chunks to process at a time for distance computation.

metric Literal['euclidean', 'cosine'] (default: 'euclidean')

The distance metric to use. The distance function can be ‘euclidean’ (default) or ‘cosine’.

between_cluster_distances Literal['nearest', 'mean_other', 'furthest'] (default: 'nearest')

Method for computing inter-cluster distances. - ‘nearest’: Standard silhouette (distance to nearest cluster) - ‘mean_other’: BRAS-specific (mean distance to all other clusters) - ‘furthest’: BRAS-specific (distance to furthest cluster)

Return type:

float

Returns:

silhouette score