netmap.downstream.clustering.unify_group_labelling

netmap.downstream.clustering.unify_group_labelling(adata, grn_adata, col_adata, col_grn_adata) → float

Adjust group labelling such that grn_adata has the same group label than the corresponding column in adata based on the grn column.

Tee prrocessed GRN Anndata object wwill contian ith an additional column in the .obs slot containing a ‘_remap’ column which contains a relabelled column matching the clustering in adata

Returns a score of the matching as the cost of the matching divided by the number of cells. The score ranges between 1/n_clusters and 1, with 1 indicating an identical clustering and 1/n_clu a random matching.

Parameters:

  • adata (AnnData) – original Anndata object containing a clustering column in the obs object specified as col_adata

  • grn_adata (AnnData) – original Anndata object containing a clustering column in the obs object specified as col_grn_adata

  • col_adata_clu (str) – Name of the reference column in expression object

  • col_grn_adata (str) – Name of the novel clustering column in GRN object

Returns:

score – The score of the linear sum assignment divided by the number of cells.

Return type:

int