Path: news.mathworks.com!newsfeed-00.mathworks.com!newsfeed2.dallas1.level3.net!news.level3.com!postnews.google.com!l35g2000vba.googlegroups.com!not-for-mail
From: Rune Allnor <allnor@tele.ntnu.no>
Newsgroups: comp.soft-sys.matlab
Subject: Re: how to deal with the inversion problem of a huge sparse
Date: Sun, 27 Sep 2009 05:49:44 -0700 (PDT)
Organization: http://groups.google.com
Lines: 21
Message-ID: <524bfbe2-ff4e-4cc6-a2d0-aa05d821fbd1@l35g2000vba.googlegroups.com>
References: <h9llp6$mho$1@fred.mathworks.com> <ace8cc81-2737-4630-83b6-76c09f427a97@p23g2000vbl.googlegroups.com> 
	<h9nmfi$n2f$1@fred.mathworks.com>
NNTP-Posting-Host: 77.17.138.11
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
X-Trace: posting.google.com 1254055784 19553 127.0.0.1 (27 Sep 2009 12:49:44 GMT)
X-Complaints-To: groups-abuse@google.com
NNTP-Posting-Date: Sun, 27 Sep 2009 12:49:44 +0000 (UTC)
Complaints-To: groups-abuse@google.com
Injection-Info: l35g2000vba.googlegroups.com; posting-host=77.17.138.11; 
	posting-account=VAp5gAkAAAAmkCze5hvZtMeedpZWNthI
User-Agent: G2/1.0
X-HTTP-UserAgent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; 
	Trident/4.0; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022; 
	.NET CLR 3.0.4506.2152; .NET CLR 3.5.30729; .NET CLR 1.1.4322),gzip(gfe),gzip(gfe)
Xref: news.mathworks.com comp.soft-sys.matlab:573188


On 27 Sep, 14:40, "Bruno Luong" <b.lu...@fogale.findmycountry> wrote:
> Rune Allnor <all...@tele.ntnu.no> wrote in message <ace8cc81-2737-4630-83b6-76c09f427...@p23g2000vbl.googlegroups.com>...

> > First of all, most processes of practical interest are
> > stationary. You don't need the long-term covariances,
> > which means that you can get away with a smaller covariance
> > matrix.
>
> You seem to infer *temporal* dimension in the OP problem.

No. 'Stationarity' means that you can draw any sub-sample
from a larger sample and the statistical properties will
be the same.

> Aren't probably mistaken between correlation (time is no needed) and cross/auto-correlation ?

Time has nothing to do with the statistical properties.
Assign any physical dimension you want to the data. I still
can't see why one would need large sparse correlation matrices.

Rune