[Bioc-devel] Subsetting affybatch and exprset objects

[Rafael A. Irizarry]

that would be a possibility. but its still hard because a lot of things 
need to be changed. remember we want [] to be a closed operation so one 
would have to return an AffyBAtch with a new cdfenv created and 
stored somehwer or 
new slot saying which probes to keep. also, we could have [[]] return 
lists of instances of probeset. but this all seems like too much trobule for 
something no-one seems to be needing. im currently leaning to permitting 
subsetting of the exprs slow with the first dimension but issuing a waring 
saying how few functions will work properly unless a new cdfenv is 
created. 


On Tue, 13 Apr 2004, A.J. Rossini 
wrote:

> 
> Would it be worth distinguishing between probesets and probes?  i.e. 
> 
> afbatch[ ]  indexing genes (and hence probesets) rather than probes?
> 
> (and I can't recall what [[ is indexing, but perhaps probes if it
> isn't?)
> 
> (don't shoot me if I'm being stupid, I still havn't had coffee this
> morning). 
> 
> best,
> -tony
> 
> 
> 
> "Rafael A. Irizarry" <ririzarr at jhsph.edu> writes:
> 
> > the problem is th if you subset eset by the first dimension then you 
> > loose the position structure, and as a consequence the locations of pms
> > and mms held in the cdfenv are no longer meaningful. if there are any 
> > suggestions on what eset[i,] should return id be happy to consider them.
> >
> > On Tue, 13 Apr 2004, Crispin Miller wrote:
> >
> >> Hi,
> >> Apologies if I should have sent this to the main bioconductor list...
> >> and tell me if you'd have preferred it there! ;-)
> >> 
> >> I've noticed an issue with subsetting ExprSet and AffyBatch objects -
> >> and wondered what what peoples' thoughts were:
> >> If I do a subsetting operation of an exprSet with 6 chips in, I get
> >> this:
> >> 
> >> > eset.mas[1,]
> >> Expression Set (exprSet) with 
> >> 
> >>    1 genes
> >>    6 samples
> >> 
> >> > eset.mas[,1]
> >> ...
> >>    22283 genes
> >>   1 samples
> >> 
> >> 
> >> However, if I do the same with AffyBatch I get this:
> >> 
> >> > eset[1,]
> >> AffyBatch object
> >> ...
> >> Number of samples=1
> >> Number of genes=22283
> >> 
> >> >eset[,1]
> >> AffyBatch object
> >> Number of samples=6
> >> Number of genes=22283
> >> 
> >> 
> >> Soooo, to subset and exprSet by chip I have to type eset[,x], but for an
> >> AffyBatch I must type: eset[x,], which is particularly confusing since
> >> AffyBatch is a subtype of exprSet... Would it be possible/a good idea,
> >> to make these two incantations the same for both classes?
> >> 
> >> Crispin
> >>  
> >> --------------------------------------------------------
> >> 
> >>  
> >> This email is confidential and intended solely for the use o...{{dropped}}
> >> 
> >> _______________________________________________
> >> Bioc-devel at stat.math.ethz.ch mailing list
> >> https://www.stat.math.ethz.ch/mailman/listinfo/bioc-devel
> >>
> >
> > _______________________________________________
> > Bioc-devel at stat.math.ethz.ch mailing list
> > https://www.stat.math.ethz.ch/mailman/listinfo/bioc-devel
> >
> 
>