[BioC] decideTests

Jason Skelton jps at sanger.ac.uk
Thu Nov 4 11:45:30 CET 2004

Hi Gordon

>Well, the error message is pretty self-explanatory!  You need to give it an object of class
>"MArrayLM", i.e., you need to give it output from lmFit() and eBayes(), as in all the examples in
>the User's Guide.
>You didn't say how you created FIT and EB.  I am guessing now that you used lm.series() or
>gls.series().  I am been phasing these functions out of the user-interface for a long time -- they
>are lower-level functions that produce unclassed list objects.  There is no need for you to use
>them.  So far I have been writing functions that are tolerant of whether the input arguments are
>formally classed or not, but in future I may not do that.
>So, please, use lmFit(), not lm.series().
Sorry, but i used lmFit()

 > COR <- duplicateCorrelation(nwaMA, ndups=2, spacing=840, DESIGN)
 > COR$cor
[1] 0.3662469
 > FIT <- lmFit(nwaMA, ndups=2, spacing=840, DESIGN, 
correlation=0.3662469, method="ls")
 > class(FIT)
[1] "MArrayLM"
[1] "limma"

I've been using this for a long time now and stopped using lm.series() a 
long time ago....

I've now managed to get this to work however
I was using the wrong version of ebayes.......?
I was using ebayes and NOT eBayes (mainly from habit)
apologies for the confusion

"you need to give it output from lmFit() and eBayes()"

Well just the output from eBayes ?
As the former gives the following message:

Error in array(x, c(length(x), 1), if (!is.null(names(x))) list(names(x),  :
        attempt to set an attribute on NULL

Thanks for your help and advice



More information about the Bioconductor mailing list