Dear Kevin, You have given us so little information that you can't really expect anyone on this list to be able to tell you what your problem is. You haven't even told us exactly what the error message is. You also need to tell us your session information, sessionInfo(). Please read the posting guide on the Bioconductor website. I am guessing that the Choleski decomposition is failing because your regression problem is singular. This could possibly arise because cor$consensus=1, or perhaps because you have MA$weights which are close to zero. There's no way for us to know this. You will need to do some more exploration of your data yourself. Best wishes Gordon >Date: Mon, 29 Oct 2007 16:46:32 +0800 >From: " ?? " <zhaoye07 at="" gmail.com=""> >Subject: [BioC] About ndups error chol(V) >To: bioconductor at stat.math.ethz.ch > >Dear all, > >I'm using the oligo microarray to find expression different genes. I decided >to use limma. This array have 3 spots for each gene. >But when I input the command >"fit<-lmFit(MA,design,ndups=3,correlation=cor$consensus)", It complains an >error on chol(V). >It works if I don't use ndups, but I need this. So, can you help me to solve >this problem? > >best wishes > >kevin