Perform SVD. The input datasets are the already preprocessed (filtering) and global-scale-nomalized datasets from Seurat.
fastSVD(samples.list, nPC = 30)
| samples.list | A list of input datasets preprocessed by Seurat. |
|---|---|
| nPC | Total number of PCs to compute and store (30 by default). |
A list consists of: PCs; Loadings; Centers (centers of samples before SVD); batch.id.forPC (batch id of cells in the samples for SVD); Raw representing the input for SVD; cells are the cell names for cells that are included and used for SVD; genes are the gene names for genes that are included and used for SVD.