scSLAT.model.loaddata.load_anndata

scSLAT.model.loaddata.load_anndata(adata, feature='PCA', noise_level=0, noise_type='uniform', edge_homo_ratio=0.9, return_PCs=False)

Create 2 graphs from single anndata (only for test)

Parameters:

• adata (anndata._core.anndata.AnnData) – Anndata object

• feature (typing.Optional[str]) – feature to use to build graph and align, now support - PCA - Harmony (default)

• noise_level (typing.Optional[float]) – node noise

• noise_type (typing.Optional[str]) – type of noise, support ‘uniform’ and ‘normal’

• edge_homo_ratio (typing.Optional[float]) – ratio of edge in graph2

• return_PCs (typing.Optional[bool]) – if return adata.varm[‘PCs’] if use feature ‘PCA’ (just for benchmark)

Return type:

typing.List

Warning

This function is only for test. It generates two graphs from single anndata by data augmentation