[R] [Stat related] Understanding Portmanteau test

[Andrew Robinson]

On Sat, Nov 08, 2008 at 09:58:14PM -0800, RON70 wrote:
> 
> Still waiting for some input. Did my question void forum rule in any manner?

No, no specific rule.  It's just not a particularly easy question to
answer, because it's not clear in what context you have seen the
phrase "portmanteau test".  Any given answer might be completely
irrelevant and useless to you.  
 
> > Sorry to be off-topic. Can somebody please explain me what is Portmanteau
> > test? Why it's name is like that? When I would say, a particular test is
> > portmanteau test? I did some googling but got no satisfactory answer at
> > all. Please anybody help for understanding that?

For what it's worth, the Cambridge Dictionary of Statistics defines a
"portmanteau test" as a test for assessing the fit of models for time
series in the presence of outliers.  It provides a citation:
Computational Statistics, 1994, 9, 301-310.

I hope that helps.

Andrew

-- 
Andrew Robinson  
Department of Mathematics and Statistics            Tel: +61-3-8344-6410
University of Melbourne, VIC 3010 Australia         Fax: +61-3-8344-4599
http://www.ms.unimelb.edu.au/~andrewpr
http://blogs.mbs.edu/fishing-in-the-bay/