Benchmarking: bm

The benchmarking module contains functions to evaluate how well specific processing steps worked, for example integration.

Batch correction metrics

bm.eval_integration(adata_post, adata_pre, ...)	Evaluate the integration.
bm.kbet(adata[, batch_key, annotation_key, ...])	kBET score.
bm.silhouette_batch(adata, batch_key, ...[, ...])	Batch ASW.
bm.pcr_comparison(adata_pre, adata_post, ...)	Principal component regression score.
bm.graph_connectivity(adata, annotation_key)	Graph Connectivity.