| smooth_via_pca {sctransform} | R Documentation |
Perform PCA, identify significant dimensions, and reverse the rotation using only significant dimensions.
smooth_via_pca(x, elbow_th = 0.025, dims_use = NULL, max_pc = 100, do_plot = FALSE, scale. = FALSE)
x |
A data matrix with genes as rows and cells as columns |
elbow_th |
The fraction of PC sdev drop that is considered significant; low values will lead to more PCs being used |
dims_use |
Directly specify PCs to use, e.g. 1:10 |
max_pc |
Maximum number of PCs computed |
do_plot |
Plot PC sdev and sdev drop |
scale. |
Boolean indicating whether genes should be divided by standard deviation after centering and prior to PCA |
Smoothed data
vst_out <- vst(pbmc) y_smooth <- smooth_via_pca(vst_out$y, do_plot = TRUE)