VoyForums
[ Show ]
Support VoyForums
[ Shrink ]
VoyForums Announcement: Programming and providing support for this service has been a labor of love since 1997. We are one of the few services online who values our users' privacy, and have never sold your information. We have even fought hard to defend your privacy in legal cases; however, we've done it with almost no financial support -- paying out of pocket to continue providing the service. Due to the issues imposed on us by advertisers, we also stopped hosting most ads on the forums many years ago. We hope you appreciate our efforts.

Show your support by donating any amount. (Note: We are still technically a for-profit company, so your contribution is not tax-deductible.) PayPal Acct: Feedback:

Donate to VoyForums (PayPal):

[ Login ] [ Contact Forum Admin ] [ Main index ] [ Post a new message ] [ Search | Check update time | Archives: 1, 2, [3] ]

[ Next Thread | Previous Thread | Next Message | Previous Message ]

Date Posted: 02:59:36 12/02/12 Sun
Author: Eva Ivits
Subject: EPCA problems

Dear list,

The expected outputs from coupled EOF analysis, what is called extended
PCA in Idrisi, are two sets of spatial patterns derived from the left
and right eigenvectors (in case of SVD algorithm used), e.g.one for the
first series and one for the second series, for each computed
components. Furthermore, two sets of principal components, or expansion
coefficients, are expected, one for the left and one for the right
matrices in case of SVD, e.g. one for the first series and one for the
second series, again for each computed components.
Now, I do get the two sets of spatial patterns but ONLY ONE SET of
principal components is computed (I've found them in the *.dif file and
individually, i.e one for each component, in the *.ipf files)!
Firstly, I do not know which series (NDVI or SPI) this set of Principal
Components relates to and secondly further analyses are not possible
anymore. One analysis would be computing homogeneous or heterogeneous
correlation maps, e.g. correlating the PCs of SPI with the SPI time
series over each pixel or correlating the PCs of SPI with the NDVI time
series over each pixels, respectively. Another important analysis would
be correlating the two sets of principal component series that would
indicate the strength of the co-variance between the two series.
My guess is that either the second set of PCs is not computed OR,
hopefully, I just did not find them.

Can anybody help?

Thanks in advance,
Eva

[ Next Thread | Previous Thread | Next Message | Previous Message ]
[ Contact Forum Admin ]

Forum timezone: GMT-8
VF Version: 3.00b, ConfDB:
Before posting please read our privacy policy.
VoyForums(tm) is a Free Service from Voyager Info-Systems.
Copyright © 1998-2019 Voyager Info-Systems. All Rights Reserved.