symphonypy.tl.ingest

symphonypy.tl.ingest(adata_query: AnnData, adata_ref: AnnData, obs: str | Iterable[str] | None = None, embedding_method: str | Iterable[str] = 'umap', labeling_method: str = 'knn', neighbors_key: str | None = None, inplace: bool = True, use_rep: str | None = None, **kwargs)

Copied from https://github.com/scverse/scanpy/blob/master/scanpy/tools/_ingest.py with little change that var_names equality between adata and adata_new wouldn’t be check if needless, and additional parameter use_rep is added.

Parameters:

• adata_query (AnnData) – target AnnData object

• adata_ref (AnnData) – source AnnData object

• obs (Optional[Union[str, Iterable[str]]], optional) – which columns from adata_red.obs to transfer, defaults to None

• embedding_method (Union[str, Iterable[str]], optional) – which adata_ref’s embeddings to transfer, defaults to “umap”

• labeling_method (str, optional) – which method to use for labeling transferring, defaults to “knn”

• neighbors_key (Optional[str], optional) – which key from adata_ref.uns to use as source of neighbors info, defaults to None

• inplace (bool, optional) – if to write directly to adata_query or return an adjusted copy of adata_query, defaults to True

• use_rep (str | None, optional) – which adata_query’s representation to use for embedding mappings. If None, it will be decided depending on circumstances, defaults to None

Returns:

if inplace=False returns a copy of adata_query with additional slots, otherwise adds to adata_query.