nichepca.graph_construction.construct_multi_sample_graph

nichepca.graph_construction.construct_multi_sample_graph#

nichepca.graph_construction.construct_multi_sample_graph(adata, sample_key, knn=None, radius=None, delaunay=False, return_graph=False, keep_local_edge_index=False, verbose=True, **kwargs)#

Construct a multi-sample graph from AnnData.

Parameters:

  • adata (AnnData) – Annotated data object.

  • sample_key (str) – Key in adata.obs where sample information is stored.

  • knn (int | None (default: None)) – Number of nearest neighbors. Defaults to None.

  • radius (float | None (default: None)) – Radius for the distance graph constructor. Defaults to None.

  • delaunay (bool (default: False)) – Whether to use the delaunay graph constructor. Defaults to False.

  • return_graph (bool (default: False)) – Whether to return the graph instead of storing it in adata. Defaults to False.

  • keep_local_edge_index (bool (default: False)) – Whether to keep the local edge index. Defaults to False.

  • verbose (bool (default: True)) – Whether to print graph statistics. Defaults to True.

  • **kwargs – Additional keyword arguments passed to the graph constructor.

Returns:

edge_index, edge_weight : torch.Tensor, torch.Tensor Edge index and edge weight of the constructed graph if return_graph is True.