[GRASS-stats] Re: [GRASS-user] Testing i.pca ~ prcomp(), m.eigensystem ~ princomp()

Nikos:

> http://grass.osgeo.org/wiki/Principal_Component_Analysis

Roger:

Good, thanks. There you say that you are using "some MODIS
surface reflectance products". I guess it will be easier to
check things if the data (GRASS location) are available, so
that others can try the same calculations. Would it be possible
to make one or more test sets available, and link them from thw
Wiki?

better yet, use the standard Spearfish SPOT imagery sample dataset
for the examples. (as m.eigensystem and now i.pca help page
examples do, probably others)

is there a way to change text color in MediaWiki? it might be
easier to highlight m.eigensystem output rows by color instead
of '#'s.

idea for possible google summer of code 2009 project: overhaul i.*
modules/lib for grass7. US$4.5k bounty :slight_smile:

cheers,
Hamish

Roger:
> Good, thanks. There you say that you are using "some MODIS
> surface reflectance products". I guess it will be easier to
> check things if the data (GRASS location) are available, so
> that others can try the same calculations. Would it be possible
> to make one or more test sets available, and link them from thw
> Wiki?

Hamish:

better yet, use the standard Spearfish SPOT imagery sample dataset
for the examples. (as m.eigensystem and now i.pca help page
examples do, probably others)

I will eventually try to re-perform all test with spearfish as well.
Except if someone else, interested in PCA, is willing to do it :-p.

Sorry that I haven't done that before. I preferred to play around with
data from my projects.

Kind regards, Nikos

On Thu, Mar 12, 2009 at 8:26 AM, Hamish <hamish_b@yahoo.com> wrote:

Nikos:

> http://grass.osgeo.org/wiki/Principal_Component_Analysis

...

better yet, use the standard Spearfish SPOT imagery sample dataset
for the examples. (as m.eigensystem and now i.pca help page
examples do, probably others)

Let me suggest the Landsat imagery in the North Carolina location
which are more recent (the Spearfish *imagery* sample dataset
isn't that great).

Markus