omicverse.pp.neighbors

omicverse.pp.neighbors(adata, n_neighbors=15, n_pcs=None, use_rep=None, knn=True, random_state=0, n_jobs=None, method='umap', transformer=None, metric='euclidean', metric_kwds=mappingproxy({}), key_added=None, copy=False, **kwargs)[source]

Compute a neighborhood graph of observations [McInnes18].

The neighbor search efficiency of this heavily relies on UMAP [McInnes18], which also provides a method for estimating connectivities of data points - the connectivity of the manifold (method==’umap’). If method==’gauss’, connectivities are computed according to [Coifman05], in the adaption of [Haghverdi16].

Parameters:

  • adata (AnnData) – Annotated data matrix.

  • n_neighbors (int (default: 15)) – The size of local neighborhood (in terms of number of neighboring data points) used for manifold approximation. Larger values result in more global views of the manifold, while smaller values result in more local data being preserved. In general values should be in the range 2 to 100. If knn is True, number of nearest neighbors to be searched. If knn is False, a Gaussian kernel width is set to the distance of the n_neighbors neighbor.

  • knn (bool (default: True)) – If True, use a hard threshold to restrict the number of neighbors to n_neighbors, that is, consider a knn graph. Otherwise, use a Gaussian Kernel to assign low weights to neighbors more distant than the n_neighbors nearest neighbor.

  • random_state (int (default: 0)) – A numpy random seed.

  • n_jobs (Optional[int] (default: None)) – Number of parallel jobs used by kNN backend. Defaults to scanpy-compatible behavior.

  • method (Optional[Literal['umap', 'gauss', 'rapids']] (default: 'umap')) – Use ‘umap’ [McInnes18] or ‘gauss’ (Gauss kernel following [Coifman05] with adaptive width [Haghverdi16]) for computing connectivities. Use ‘rapids’ for the RAPIDS implementation of UMAP (experimental, GPU only). Use ‘torch’ for GPU-accelerated connectivity computation.

  • transformer (Optional[str] (default: None)) – KNN search implementation. Options: None (auto), ‘pyg’ (PyTorch Geometric, recommended for GPU), ‘pynndescent’, ‘sklearn’, or ‘rapids’. ‘pyg’ provides 20-100× speedup over other methods. In cpu-gpu-mixed mode, None defaults to ‘pyg’.

  • metric (Union[Literal['cityblock', 'cosine', 'euclidean', 'l1', 'l2', 'manhattan'], Literal['braycurtis', 'canberra', 'chebyshev', 'correlation', 'dice', 'hamming', 'jaccard', 'kulsinski', 'mahalanobis', 'minkowski', 'rogerstanimoto', 'russellrao', 'seuclidean', 'sokalmichener', 'sokalsneath', 'sqeuclidean', 'yule'], Callable[[ndarray, ndarray], float]] (default: 'euclidean')) – A known metric’s name or a callable that returns a distance.

  • metric_kwds (Mapping[str, Any] (default: mappingproxy({}))) – Options for the metric.

  • key_added (Optional[str] (default: None)) – If not specified, the neighbors data is stored in .uns[‘neighbors’], distances and connectivities are stored in .obsp[‘distances’] and .obsp[‘connectivities’] respectively. If specified, the neighbors data is added to .uns[key_added], distances are stored in .obsp[key_added+’_distances’] and connectivities in .obsp[key_added+’_connectivities’].

  • copy (bool (default: False)) – Return a copy instead of writing to adata.

Return type:

Optional[AnnData]

Returns: Depending on copy, updates or returns adata with the following:

See key_added parameter description for the storage path of connectivities and distances.

connectivitiessparse matrix of dtype float32.

Weighted adjacency matrix of the neighborhood graph of data points. Weights should be interpreted as connectivities.

distancessparse matrix of dtype float32.

Instead of decaying weights, this stores distances for each pair of neighbors.

Notes

If method=’umap’, it’s highly recommended to install pynndescent pip install pynndescent. Installing pynndescent can significantly increase performance, and in later versions it will become a hard dependency.