[Rd] Bug in getVarCov.gls method (PR#9752)

[ripley at stats.ox.ac.uk]

Thank you, will be fixed in the next release of nlme (not yet scheduled, 
as everythign is code frozen for the release of R 2.5.1 tomorrow).

On Tue, 26 Jun 2007, Andrzej Galecki wrote:

>
> Two attachments:
>
> 1. getVarCovBugReport.R  - Rcode with an example illustrating the problem and 
> how to fix it
> 2. getVarCovBugReportSession.txt contains code and R session results.
>
> Thank you
>
> Andrzej Galecki
>
>
>
> Prof Brian Ripley wrote:
>
>> On Mon, 25 Jun 2007, agalecki at umich.edu wrote:
>> 
>>> I am using R2.5 under Windows.
>> 
>> 
>> I presume you mean 2.5.0 (there is no R2.5: see the posting guide).  But 
>> which version of nlme, which is the relevant fact here?  The R posting 
>> guide suggests showing the output of sessionInfo() to establish the 
>> environment used.
>> 
>>> Looks like the following statement
>>> 
>>> vars  <-  (obj$sigma^2)*vw
>>> 
>>> in getVarCov.gls method (nlme package) needs to  be replaced with:
>>> 
>>> vars <- (obj$sigma*vw)^2
>> 
>> 
>> We need some evidence!  Please supply a reproducible example and give your 
>> reasoning.  For example, the FAQ says
>>
>>   The most important principle in reporting a bug is to report _facts_,
>>   not hypotheses or categorizations.  It is always easier to report the
>>   facts, but people seem to prefer to strain to posit explanations and
>>   report them instead.  If the explanations are based on guesses about how
>>   R is implemented, they will be useless; others will have to try to figure
>>   out what the facts must have been to lead to such speculations.
>>   Sometimes this is impossible.  But in any case, it is unnecessary work
>>   for the ones trying to fix the problem.
>>
>>   It is very useful to try and find simple examples that produce
>>   apparently the same bug, and somewhat useful to find simple examples
>>   that might be expected to produce the bug but actually do not.  If you
>>   want to debug the problem and find exactly what caused it, that is
>>   wonderful.  You should still report the facts as well as any
>>   explanations or solutions. Please include an example that reproduces the
>>   problem, preferably the simplest one you have found.
>> 
>> It should be easily possible to cross-check an example with one of the many 
>> other ways available to do GLS fits in R.
>> 
>> [...]
>> 
>> 
>
>

-- 
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