[BioC] limma object column subsetting

[Gordon K Smyth]

> Date: Tue, 14 Dec 2004 13:42:48 -0500
> From: Matt McCall <mmccall at mail.nih.gov>
> Subject: [BioC] limma object column subsetting
> To: bioconductor at stat.math.ethz.ch
>
> In the old limma, one could subset a MAlist or RGlist (let's say RGnorm
> is a MAlist) by entering:
> RGnorm[,1:3] for the first three columns for example
> In the version, I'm using now (v1.8.1) this no longer works.

It does still work as far as I can see.

>  > RGnorm[,1]
> Error in object$targets[j, , drop = FALSE] :
> 	incorrect number of dimensions
>
> Can someone help me with this. I'm running R v2.0.0

You don't make it clear what classs of object your RGnorm is.  I am assuming it is an RGList or an
MAlist.  The problem seems to be that the 'targets' component of your object is a simple vector
rather than a data frame as it is expected to be.

How did you create the RGnorm and, in particular, how did you set the targets component?  The only
limma function which creats an RGList or MAList with the targets component set automatically is
read.maimages(), and in that the case the 'targets' should be subsettable correctly.

However you created RGnorm, you should be able to fix the above problem by

RGnorm$targets <- as.data.frame(RGnorm$targets)

or by

RGnorm$targets <- NULL

Gordon

>           _
> platform powerpc-apple-darwin6.8
> arch     powerpc
> os       darwin6.8
> system   powerpc, darwin6.8
> status
> major    2
> minor    0.0
> year     2004
> month    10
> day      04
> language R
>
> Matt McCall