Re: [aroma.affymetrix] mufColumns on genes with more than 500 probes across the gene

2015-06-05 Thread Mark Robinson
Hi Iain, Can you send me a reproducible example, plus the output of sessionInfo() and traceback() after the error, please? It'll probably be too large for email, so maybe use Dropbox or dropsend .. My guess is that the units with 500 probes actually have NAs. Maybe you can check this also

[aroma.affymetrix] mufColumns on genes with more than 500 probes across the gene

2015-06-04 Thread Iain
Hi We're trying to calculate the difference in residuals between two groups and generate mufScores (FIRMAGene) on those residual differences. Calculating the differences goes smoothly: rsu_diff - lapply(unlist(rsu, use.names=FALSE), byrow=FALSE, ncol=length(unique(cls))) where cls is a