Skip to content

bug in functions of form "mcpar<-"

6 messages · Peter Dalgaard, Paul Gilbert, Martyn Plummer

Original
Martyn Plummer
# 
This fairly harmless looking piece of code, which worked in
0.61 (and in S-PLUS) fails  in 0.62.1. 

R> "mcpar<-" <- function(x,mcpar) {attr(x,"mcpar") <- mcpar; x}
R> mcpar(x) <- c(1,100,1)
Error in mcpar<-(*tmp*, value = c(1, 100, 1)) : unused argument to function

The error message gives a hint about how to work around the problem -
just replace the argument "mcpar" with "value". 

R> "mcpar<-" <- function(x,value) {attr(x,"mcpar") <- value; x}
R> mcpar(x) <- c(1,100,1)
R> 

This breaks release 0.3-3 of my "coda" package, but I will fix it.
Martyn
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Peter Dalgaard
# 
Martyn Plummer <plummer@iarc.fr> writes:
Hmm. We had some pretty antisocial consequences from *not* using the
value= form with John Chambers' [<-.data.frame code, which assumes
that it is used that way in order to see whether i,j, or both are
absent. 

So Splus is using value= on indexing operations only?
Paul Gilbert
# 
I'm having the same problem without a clue to a work around. (Mind you, the work
around for mcpar wasn't so obvious to me.)

Paul Gilbert
____

R : Copyright 1998, The R Development Core Team
Version 0.62.1  (June 15, 1998)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type    "?license" or "?licence" for distribution details.

R is a collaborative project with many contributors.
Type    "?contributors" for a list.

Type    "demo()" for some demos, "help()" for on-line help, or
        "help.start()" for a HTML browser interface to help.
[1] ".GlobalEnv"   "Autoloads"    "package:base"
{
 attr(x, "tframe") <- tf
 x
}

+ + + + > >
Error in tframe<-(*tmp*, value = c(1, 1000, 1)) : unused argument to function
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Paul Gilbert
# 
The light came on. I was thinking the problem with mcpar was that the argument
had the same name as the function, but it is that the argument must be called
"value".  If R is intended to be like this it seems like a pretty serious
incompatibility with S (not to mention that I will have to make a lot of
changes).

Paul Gilbert

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Martyn Plummer
#
Peter Dalgaard BSA wrote:

            

      
      
      
    

        
Looks like it.  As I said, my other example works in S-PLUS, but try
doing
the same thing with an assignment on an index and you get an error:
(This examples is obviously silly in order to distinguish it from the
default
indexing function).

S> "[<-.foo" <-
function(x,i,value){y<-unclass(x);y[i]<-value+1;class(y)<-"foo";y}
S> x <- 1:10
S> class(x) <- "foo"
S> x[1]<-100
S> x
 [1] 101   2   3   4   5   6   7   8   9  10
attr(, "class"):
[1] "foo"
S> "[<-.foo" <-
function(x,i,bar){y<-unclass(x);y[i]<-bar+1;class(y)<-"foo";y}
S> x[1]<-200
Error in call to [<-.foo(): Argument value= not matched
Dumped

Martyn
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

  
  


  


      

    

    

      
      

        


  

        

      Paul Gilbert
    

    

      
      #
    

  


  

          

      
      
      
    

        
Definitely Splus does not in general require "value" as the name of the
argument. I've changed about a dozen instances of this in my code. It turned out
not to be as bad a job as I expected, debugging because of the one change I
missed was the worst part. But everything was working fine in Splus.

Paul Gilbert

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._