[BioC] limma 1.3.4 for Win

[Gordon Smyth]

limma for R 1.8.1 is available from http://bioinf.wehi.edu.au/limma. The 
latest version is 1.3.18.

Unfortunately I don't think I can put these versions on the BioC site. BioC 
release policy does not allow updates to the official 1.2 release version, 
except to fix minor and clear bugs, and the development version must run 
with R 1.9.0.

See https://stat.ethz.ch/pipermail/bioconductor/2003-December/003398.html

Gordon

At 10:03 PM 19/02/2004, Stephen Henderson wrote:
>Hello
>I am having some problems with limma again. specifically when using this for
>affy data. This works fine on my laptop and Linux computer but not on my
>colleagues computer where I have tried to install this for him
>
>His setup is the current limma version 1.3.1 running on R1.8.1 and Windows
>XP. I get:
>
> > contrast.matrix<-makeContrasts(group2-group1, group3-group2,levels=design)
> > fit2<-contrasts.fit(fit1,contrast.matrix)
> > fit3<-eBayes(fit2)
>Error in ebayes(fit = fit, proportion = proportion, std.coef = std.coef) :
>         No residual degrees of freedom in linear model fits
>
>Somehow I have limma 1.3.4 on my laptop!!! which works well but 1.3.1 is on
>the site, 1.5 is in the devel folder but only works with R1.9 (not really
>released yet!).
>
>SO...Can we have limma 1.3.4 again please??
>
>Thanks
>Stephen Henderson
>WIBR, UCL