[Rd] issue with [[<-Call

[Prof Brian Ripley]

The issue is that you are using [["weights"]] to add an element.  It works 
for existing eleents as well as for [<- and $<-.

However, I would have done this without a loop

 	nm <- c("data", "weights", "subset", "na.action")
 	extras <- nm %in% names(Call)
 	temp[ nm[extras] ] <- Call[extras]

which will also handle adding NULL elements correctly.

And BTW, the (current) R idiom is to use eval.parent(), and lm() shows 
another way to do this that I would see as cleaner.


On Thu, 25 Dec 2008, Charles C. Berry wrote:

> On Thu, 25 Dec 2008, Terry Therneau wrote:
>
>>  The following code works in Splus but not in R
>> 
>> coxph <- function(formula, data, weights, subset, na.action,
>> 	init, control, method= c("efron", "breslow", "exact"),
>> 	singular.ok =TRUE, robust=FALSE,
>> 	model=FALSE, x=FALSE, y=TRUE, ...) {
>>
>>    method <- match.arg(method)
>>    Call <- match.call()
>>
>>    # create a call to model.frame() that contains the formula (required)
>>    #  and any other of the relevant optional arguments
>>    # then evaluate it in the proper frame
>>    temp <- call('model.frame', formula=formula)
>>    for (i in c("data", "weights", "subset", "na.action")) #add optional 
>> args
>>        if (!is.null(Call[[i]])) temp[[i]] <- Call[[i]]
>>
>>    if (is.R()) m <- eval(temp, parent.frame())
>>    else        m <- eval(temp, sys.parent())
>> 
>> ---------------------------------
>> 
>> The problem is that the names ('data', 'weights', etc) do no propogate over
>> to temp, the new call object.  It looks like an oversight in the 
>> replacement
>> method.
>
>
>       if (!is.null(Call[[i]])) temp[ i ] <- Call[[i]]
>
> should do it in R at least.
>
> HTH,
>
> Chuck
>
>
>>
>>  Priority: low.  I have other code forms that work in both dialets.
>>
>>  I created this when teaching an internal class on S programming; the goal
>> was to make what was happening as transparent as possible.  If anyone has
>> something that they think is even better from a teaching point of view I'd
>> be delighted to see it.
>>
>> 	Terry T.
>> 
>> ______________________________________________
>> R-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel
>> 
>
> Charles C. Berry                            (858) 534-2098
>                                            Dept of Family/Preventive 
> Medicine
> E mailto:cberry at tajo.ucsd.edu	            UC San Diego
> http://famprevmed.ucsd.edu/faculty/cberry/  La Jolla, San Diego 92093-0901
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595