performance of apply

[Ross Ihaka]

	Subject: Re: performance of apply
	From: Peter Dalgaard BSA <p.dalgaard@biostat.ku.dk>
	Date: 29 May 1998 18:04:30 +0200
	MIME-Version: 1.0
	Content-Transfer-Encoding: 7bit

	Oh, dear. Looking at apply, we have right in the middle of it:

	            for (i in 1:d2) ans[[i]] <- FUN(array(newX[, i], d.call, 
	                    dn.call), ...)

	Assignment to the tail of a very long list... 

	I think I'm beginning to understand why Ross keeps talking about
	generic vectors. Either that or we need to add some low-level list
	primitives to the language. 


Got it in one.  To reach ans[[i]] you have to cdr down the whole list.
The subsetting code would be vastly simplified without the special
list code.  Ugliest of all is subscript manipulation for lists with
dim attributes.

[ I estimate 2 months for a fix - not that it will take that long,
  but there are some other things higher up the todo list ].

	Ross
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._