scSLAT.model.utils.add_noise
- scSLAT.model.utils.add_noise(adata, noise='nb', inverse_noise=5)[source]
Add poisson / negative binomial noise on raw counts also run scanpy pipeline to PCA step
- Parameters:
adata – anndata object
noise (
typing.Optional[str]) – type of noise, one of ‘poisson’ or ‘nb’inverse_noise (
typing.Optional[float]) – if noise is ‘nb’, control the noise level (smaller means larger variance)
- Return type: