Re: EOF explained variance
- From: "Roger Stafford" <ellieandrogerxyzzy@xxxxxxxxxxxxxxxxxxxxxx>
- Date: Mon, 14 Jul 2008 16:32:02 +0000 (UTC)
"Leandro Calil" <leandro@xxxxxxxxxxxxxxxx> wrote in message <g5fn4v$nj7
$1@xxxxxxxxxxxxxxxxxx>...
Hi guys
I am a intermediate user of matlab and I need your help...
I try to calculate the explained variance for eof analysis,
but when I try to calculate the covariance matrix from my
data, matlab shown "out of memory". My data size is 126x8036
and my computer have 1 Gb ram. I try Cov(X) and a simple
X*X'... I don?t known another way to calculate covariance
matrix that use less memory. If some body can help me??
thanks a lot...
Leandro
The documentation for 'cov' states that "each row is an observation, and
each column is a variable". If your X matrix is of size 126 x 8036, then cov(X)
will assume you have 8036 variables with 126 observations and produce a
covariance matrix of size 8036 x 8036, which is very large. Is that what you
intended? The matrix product you wrote, X*X', is only 126 x 126 in size.
This suggests that perhaps you actually have 8036 observations and only 126
variables and need to compute cov(X').
Roger Stafford
.
- Follow-Ups:
- Re: EOF explained variance
- From: Leandro Calil
- Re: EOF explained variance
- References:
- EOF explained variance
- From: Leandro Calil
- EOF explained variance
- Prev by Date: Need a more detail shapefile for Quebec Canada
- Next by Date: Re: gui problems
- Previous by thread: EOF explained variance
- Next by thread: Re: EOF explained variance
- Index(es):
Relevant Pages
|
