Giters

kkdey / CorShrink

R package for adaptive correlation and covariance matrix shrinkage.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Question about options

kayserb opened this issue · comments

commented

Hi Kushal - This is a really great method. I had a couple questions regarding the different options available:

  1. For Spearman correlations, can I input the rho and n into the CorShrinkVector() function, or do rank correlations require bootstrapped-derived SEs for valid inference?

  2. I have a large number of correlations, but I am only interested in positive correlations. This will also save computational time because it is A LOT of correlations. If I subset to only positive correlations and feed them into CorShrinkVector(), would specifying the "+uniform" distribution be the best way to proceed?