Thanks to Robert and Gabor for their replies, but neither was what I was
looking for, undoubtedly because of the poor phrasing of my question (+ a
typo -- however, even if correctly typed it doesn't work). I finally
realized that the "elegant" approach I sought can easily be done without
setGeneric:
setMethod('[','foo',function(x,i,j,drop,...){
cl<-as.list(sys.call())
cl[[2]]<[EMAIL PROTECTED]
eval(as.call(cl))
})
Indeed, this is a general template for this sort of thing. Should have
thought of this before, as V&R's S PROGRAMMING has numerous such examples...
Sigh...
-- Bert
-----Original Message-----
From: Gabor Grothendieck [mailto:[EMAIL PROTECTED]
Sent: Monday, June 20, 2005 1:31 PM
To: Berton Gunter
Cc: [email protected]
Subject: Re: [R] How to define S4 methods for '['
On 6/20/05, Berton Gunter <[EMAIL PROTECTED]> wrote:
> Folks:
>
> This is a question about the S4 formal class system.
>
> Suppose I have a class, 'foo', defined by:
>
> setClass('foo',representation(dat='matrix', id='character') )
>
> I wish to define a '[' method for foo that will extract from the 'dat'
slot.
> I would have thought that the following would work, but it doesn't:
>
> setMethod("[","foo",function(x,i, j, .,drop=TRUE)callGeneric([EMAIL
> PROTECTED],i,
> j,drop=drop) )
>
> The only way I have succeeded in defining this method is using brute force
> eval(parse(. :
>
> {eval(parse(text=paste('.dat(x)[',
> ifelse(missing(i),',','i,'),
> ifelse(missing(j),']','j]'))))
> }
>
> This works. However, I am not able under any circumstances to pass the
drop
> argument -- it is ignored.
>
> I would appreciate any pointers about how to do this properly. If this is
> explicitly in the Green Book (I do not have it with me at the moment),
that
> will suffice.
>
Download the source to the 'its' package where an S4 [ method is defined.
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
