stereo.core.StPipeline.pca

StPipeline.pca(use_highly_genes=False, n_pcs=None, svd_solver='auto', hvg_res_key='highly_variable_genes', random_state=0, dtype='float32', res_key='pca')

Principal component analysis.

Parameters:

• use_highly_genes (bool) – whether to use the expression of hypervariable genes only.

• n_pcs (Optional[int]) – the number of principle components to compute.

• svd_solver (Literal['auto', 'full', 'arpack', 'randomized']) –

  default to 'auto'.

  • If 'auto' :

    The solver is selected by a default policy based on X.shape and n_pcs: if the input data is larger than 500x500 and the number of components to extract is lower than 80% of the smallest dimension of the data, then the more efficient ‘randomized’ method is enabled. Otherwise the exact full SVD is computed and optionally truncated afterwards.

  • If 'full' :

    run exact full SVD calling the standard LAPACK solver via scipy.linalg.svd and select the components by postprocessing

  • If 'arpack' :

    run SVD truncated to n_pcs calling ARPACK solver via scipy.sparse.linalg.svds. It requires strictly 0 < n_pcs < min(x.shape)

  • If 'randomized' :

    run randomized SVD.

• hvg_res_key (Optional[str]) – the key of highly variable genes to get targeted result,`use_highly_genes=True` is a necessary prerequisite.

• random_state (Union[None, int, RandomState]) – change to use different initial states for the optimization, fixed value to fixed result.

• dtype (str) – numpy data type string to which to convert the result.

• res_key (str) – the key for storage of PCA result.

Returns:

Computation result of principal component analysis is stored in self.result where the result key is 'pca'.