[Rd] Re: [R] H-F corr.: covariance matrix for interaction effect

Bela Bauer bela_b at gmx.net
Tue Mar 1 12:27:36 CET 2005

Peter Dalgaard wrote:
> Just in case anyone was wondering, I think I now know what SAS is
> doing, and yes, it is a bug. 

I was still wondering. :-)
Now, this is quite some interesting news, since everyone here has been 
relying on SAS H-F corrections, and I don't think they'll be amused to 
know that they might've been wrong.

I've been glancing over to Rd, but I didn't see anything about it - do 
you have any code yet that I could look at to figure out what you do? It 
doesn't need to be complete yet, just something that will tell me about 
the algorithm...



More information about the R-devel mailing list