scanpy.pp.neighbors
- scanpy.pp.neighbors(adata, n_neighbors=15, n_pcs=None, use_rep=None, knn=True, random_state=0, method='umap', metric='euclidean', metric_kwds=mappingproxy({}), key_added=None, copy=False)
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'
). Ifmethod=='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
isTrue
, number of nearest neighbors to be searched. Ifknn
isFalse
, a Gaussian kernel width is set to the distance of then_neighbors
neighbor.- n_pcs :
Optional
[int
] (default:None
) Use this many PCs. If
n_pcs==0
use.X
ifuse_rep is None
.- use_rep :
Optional
[str
] (default:None
) Use the indicated representation.
'X'
or any key for.obsm
is valid. IfNone
, the representation is chosen automatically: For.n_vars
< 50,.X
is used, otherwise ‘X_pca’ is used. If ‘X_pca’ is not present, it’s computed with default parameters.- knn :
bool
(default:True
) If
True
, use a hard threshold to restrict the number of neighbors ton_neighbors
, that is, consider a knn graph. Otherwise, use a Gaussian Kernel to assign low weights to neighbors more distant than then_neighbors
nearest neighbor.- random_state :
Union
[None
,int
,RandomState
] (default:0
) A numpy random seed.
- 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).
- 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.
- adata :
- Return type
- Returns
Depending on
copy
, updates or returnsadata
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.
- connectivitiessparse matrix of dtype
Notes
If
method='umap'
, it’s highly recommended to install pynndescentpip install pynndescent
. Installingpynndescent
can significantly increase performance, and in later versions it will become a hard dependency.