napari_prism.tl.cluster_scores

Contents

napari_prism.tl.cluster_scores#

napari_prism.tl.cluster_scores(adata, clustering_score, k=None, inplace=False)#

Assess the quality of clustering results by computing a pairwise clustering score between every other clustering run. Higher values usually indicate concordant clustering results.

Parameters:

  • adata (AnnData) – Anndata object containing clustering results.

  • clustering_score (Literal['ARI', 'NMI', 'AMI']) – The clustering score to assess. Scores either adjusted rand index (ARI), normalized mutual information (NMI), or adjusted mutual information (AMI).

  • k (Optional[int] (default: None)) – Subset the pairwise scores to a given k.

Returns:

pandas DataFrame of pairwise scores between clustering runs.

Contents