search

yusuke-imoto-lab / RECODE

RECODE - resolution of curse of dimensionality in single-cell data analysis
https://yusuke-imoto-lab.github.io/RECODE/
14 stars 2 forks source link

Fast PCA in R #1

Open tomsing1 opened 2 years ago

tomsing1 commented 2 years ago

You mentioned in the README that the R implementation is slow due to the use of the prcomp function. Have you taken a look at the irlba package, yet? It offers several very fast algorithms for

truncated singular value decomposition and principal components analysis of large sparse and dense matrices.

The BiocSingular package is another great resource, it also offers parallelization e.g. across multiple cores.

yusukeimoto commented 1 year ago

Thank you for your comment. I tested both regular pca and prcomp functions, but their speeds do not significantly differ in my environment.