R devel - Nov 1997 - R-alpha: Re: R/R-minus incompatibility [ an old one ]

>>>>> "FrL" == Friedrich Leisch
<Friedrich.Leisch@ci.tuwien.ac.at> writes:
    FrL> Hmm, I'm quite perplexed by something very
trivial/stupid/whatever
    FrL> ...  don't know, if this one has been reported before:

    R> x<-1 x[1:2]
    FrL> Error: subscript out of bounds

    Splus> x<-1 x[1:2]
    FrL> [1] 1 NA

    FrL> Or is there an option to control this behaviour?

Fritz, this *is* known and is on purpose.
However, it seems that it is  NOT in the R-FAQ  (Kurt ?!?)

-- Martin

--------------- Here are two example mails from R-devel ------------------


 From: Peter Dalgaard BSA <p.dalgaard@kubism.ku.dk>
 Date: 25 Apr 1997 01:36:23 +0200
 To: r-devel@stat.math.ethz.ch
 Subject: R-alpha: list assignment

 This works in Splus:

 > x<-list()
 > x[["f"]]<-1
 > zz<-"g"
 > x[[zz]]<-2
 > x
 $f:
 [1] 1

 $g:
 [1] 2

 In R both variants fail unless the name is already on the list. The
 first one can be replaced by x$f, but there's seems to be no
 substitute for the other one (oh yes I found one, but it's not fit to
 print!). This comes up if you e.g. want to create a variable in a data
 frame with a name given by a character string.

------------------------------------------------------

 From: Robert Gentleman <rgentlem@stat.auckland.ac.nz>
 Date: Fri, 25 Apr 1997 13:12:36 +1200 (NZST)
 To: r-devel@stat.math.ethz.ch
 Subject: Re: R-alpha: list assignment

 Peter,
   They don't work on purpose. Basically this is the same problem as
indexing
 beyond the end of an array. In almost all cases it is an error.
 To add a variable with a specific name to a data frame how about

 x (the dataframe)
 var1 the variable
 name1 the name

  x<-cbind(x,var1)
  names(x)[ncol(x)]<-name1

 robert
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-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
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

>> x[["f"]]<-1
 >> zz<-"g"
 >In R both variants fail unless the name is already on the list. The
 >first one can be replaced by x$f, but there's seems to be no
 >substitute for the other one (oh yes I found one, but it's not fit to

A long time ago it seems it was decided to remain incompatible with S on
this, I think because  people often do this in error. (Indeed, it did
point out some errors in my code.) However, there were some instances in
my code where I did something like
x[[c("f","g","h")]] <- 1:3 and wanted
the corresponding elements updated or inserted as appropriate.  I have
replaced those instances with listadd(x,
c("f","g","h")) <- 1:3 where
listadd is as defined below (more or less as suggested by Ross or Robert
at the time). It is one of the functions in the compatibility code I
have circulated previously.

Paul Gilbert
________
   "list.add<-" <- function(x, replace, value)
     {# replace or add elements to a list.
      if (is.numeric(replace))
        {x[[replace]] <- value
         return(x)
        }
      if (is.null(value))  value <- list(NULL)
      if (!is.list(value)) value <- list(value)
      if (1 == length(value))
       {for (i in seq(length(replace)))
          x<- do.call("$<-", list(x,replace[i],value[[1]]))
       }
      else
        {if(length(value) != length(replace) )
         stop("number of replacement values != number of elements to
replace")
         for (i in seq(length(replace)))
            x<- do.call("$<-", list(x,replace[i],value[[i]]))
        }
      x
     }


=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-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
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=