R help - Oct 2004 - Evaluating Assignment-Operator R 2.0.0

Dear Helpers,

in which way I have to use the assignment-operator evaluating left to right 
with R 2.0.0?

With R 2.0.0 (w2k-installation) the "superassignment" operator
'<<-' seems to
have a different behaviour compared with R 1.9.1 :

  > x<<-4
  > x
  [1] 4
  > attr(x,'y')<<-5
  Error: Object "x" not found
  >

Using earlier versions of R I never got this error.

In the NEWS file and the docs coming along with R 2.0.0 I did not found 
anything concerning this problem.
AFAIK '=' instead of '<<-' works - but is allowed only for
toplevel-
assignments. 

Thanks in advance, Friedemann

On Fri, 8 Oct 2004, friedemann wrote:
> Dear Helpers,
> 
> in which way I have to use the assignment-operator evaluating left to right
> with R 2.0.0?
<- is the operator you should normally use.
> With R 2.0.0 (w2k-installation) the "superassignment" operator
'<<-' seems to
> have a different behaviour compared with R 1.9.1 :
> 
>   > x<<-4
>   > x
>   [1] 4
>   > attr(x,'y')<<-5
>   Error: Object "x" not found
>   >
> 
> Using earlier versions of R I never got this error.
> 
> In the NEWS file and the docs coming along with R 2.0.0 I did not found 
> anything concerning this problem.
Did you not notice NEWS contains

   o   Complex superassignments were wrong when a variable with the same
        name existed locally, and were not documented in R-lang.
> AFAIK '=' instead of '<<-' works - but is allowed
only for toplevel-
> assignments. 
= is allowed almost everywhere <- is, and not just at top-level.
I don't know what you are really trying to do, but

x <- 4
testit <- function() attr(x,'y') <<- 5
testit()
x

works as documented, and that is the normal type of usage of
superassignments.  As the NEWS entry says, R-lang (`The R Language 
Definition') gives fuller details (and I guess the help page for <<- 
should point there).

-- 
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

"friedemann" <friedemann at gehrt.info> writes:
> With R 2.0.0 (w2k-installation) the "superassignment" operator
'<<-' seems to
> have a different behaviour compared with R 1.9.1 :
> 
>   > x<<-4
>   > x
>   [1] 4
>   > attr(x,'y')<<-5
>   Error: Object "x" not found
>   >
> 
> Using earlier versions of R I never got this error.
> 
> In the NEWS file and the docs coming along with R 2.0.0 I did not found 
> anything concerning this problem.
It's there. Look for "complex superassignment".
> AFAIK '=' instead of '<<-' works - but is allowed
only for toplevel-
> assignments. 
Eh? You mean "assignments in the current environment"? 

Why do you want to use <<- at top level (the command-line) anyway? 

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907