R help - Mar 2004 - update/offset

I've noticed that the update() function does not seem to work correctly
when offset(..) terms are there:

update(modelwithoffset,  .~.-afixedeffect)

drops the offset term.

I'm using this with a negbin model, but I think it goes wrong with the
update.formula() function.

update.formula
function (old, new, ...)
{
    env <- environment(as.formula(old))
    tmp <- .Internal(update.formula(as.formula(old), as.formula(new)))
    out <- formula(terms.formula(tmp, simplify = TRUE))
    environment(out) <- env
    return(out)
}
<environment: namespace:base>



The .Internal command in the code is where my efforts grind to a halt.


Any suggestions (apart from redefine the new model from scratch) ?

Thanks

Simon Bond.

Which version of R?  This is supposedly fixed in 1.9.0 beta, so please try 
that.

On Fri, 26 Mar 2004, Simon.Bond wrote:
> I've noticed that the update() function does not seem to work correctly
> when offset(..) terms are there:
> 
> update(modelwithoffset,  .~.-afixedeffect)
> 
> drops the offset term.
> 
> I'm using this with a negbin model, but I think it goes wrong with the
> update.formula() function.
> 
> update.formula
> function (old, new, ...)
> {
>     env <- environment(as.formula(old))
>     tmp <- .Internal(update.formula(as.formula(old), as.formula(new)))
>     out <- formula(terms.formula(tmp, simplify = TRUE))
>     environment(out) <- env
>     return(out)
> }
> <environment: namespace:base>
> 
> 
> 
> The .Internal command in the code is where my efforts grind to a halt.
The prohlem was in terms.formula(tmp, simplify = TRUE).
> Any suggestions (apart from redefine the new model from scratch) ?

-- 
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