rapids_singlecell.pp.regress_out#
- rapids_singlecell.pp.regress_out(adata, keys, *, layer=None, inplace=True, batchsize=100, verbose=False)[source]#
Use linear regression to adjust for the effects of unwanted noise and variation.
- Parameters:
- adata AnnData
AnnData object
- keys str | list
Keys for numerical observation annotation on which to regress on.
- layer str | None (default:
None
) Layer to regress instead of
X
. IfNone
,X
is regressed.- inplace bool (default:
True
) Whether to update
adata
or return the corrected matrix ofadata.X
andadata.layers
.- batchsize int | Literal[‘all’] | None (default:
100
) Number of genes that should be processed together. If
'all'
all genes will be processed together if.n_obs
<100000. IfNone
each gene will be analysed separately. Will be ignored if cuML version < 22.12- verbose bool (default:
False
) Print debugging information
- Return type:
cp.ndarray | None
- Returns:
Returns a corrected copy or updates
adata
with a corrected version of the originaladata.X
andadata.layers['layer']
, depending oninplace
.