On Sun, 2009-03-22 at 23:53 +0100, Nikos Alexandris wrote:
I have prepaired a grass-location
containing the 3 modis bands I've used for all PCA related examples. I
will eventually upload them in gregis [1]
If anyone would like to try what I tried here are the MODIS bands:
# grass location
Nikos:
If outputs are not identical, either R or grass do some hidden
modification or there is a bug in either grass or R (all within
limits, e.g. identical up to the 5th digit in scientific format is
fine?).
Some textbooks give a rule of thumb for further analysis to use only
than 1. The reasoning is fairly weak, but goes like this: if a PC has
eigenvalue 1, it explains more variance than any of the original
variables, which all have variance 1.
Maybe I should Cc: this to the wiki.
--
Edzer
Or even better include it in the docs if there is anything in your
Markus Neteler wrote:
Remember that you have to rescale these data first:
https://lpdaac.usgs.gov/lpdaac/products/modis_products_table
- MOD09Q1 Terra
Surface Reflectance Bands 1–2
Tile 250m
8 Day
Edzer Pebesma wrote:
Markus Metz wrote:
I'm more familiar with non-spatial PCA, so it's high time I read the
manual of i.pca, and the new wiki page on it...
I think there's no such thing as spatial or non-spatial PCA. There's
just PCA.
That was a feeble attempt to buy time
Edzer Pebesma wrote:
Markus, a few notes:
- if you do PCA on uncentered data, by computing the eigenvalues of the
uncentered covariance matrix, this implies that bands with a larger mean
will get more influence on the final PCAs. I have sofar not managed
finding an argument why this would be
Markus:
It seems that i.pca output is supposed to be identical to
prcomp(center=FALSE, scale=FALSE) output in R, because a PCA is
scale-sensitive and the eigenvalue as reported by i.pca is the variance
of the raw, unstandardised data.
The thing is that with the SPOT data all seems fine and
On Wed, 2009-04-01 at 18:21 +0200, Edzer Pebesma wrote:
Markus, a few notes:
- if you do PCA on uncentered data, by computing the eigenvalues of the
uncentered covariance matrix, this implies that bands with a larger mean
will get more influence on the final PCAs. I have sofar not managed
Nikos,
I'm probably missing a message (I do not see which is
exactly the difference between the spot and the modis
example), but
Perhaps you must use
r.mapcalc to rescale and not r.rescale, or if using r.rescale you must
use r.mapcalc afterwards to rewrite the output raster, as r.rescale
is
Agustin:
Nikos, I'm probably missing a message (I do not see which is
exactly the difference between the spot and the modis example),
Hi Agus. The difference is:
Using SPOT (range up to max. 255):
i.pca == prcomp(x, center=TRUE, scale=FALSE)
Using MODIS (range varries between bands, up to
On Tue, 2009-03-31 at 12:56 +0200, Edzer Pebesma wrote:
I tried to improve the wiki page, but never got access.
--
Edzer
Edzer, it definitely needs to be improved. Hopefully you'll get access
while the thing is boiling.
@admins: Please, we need access ;-)
Kindest regards, Nikos
Nikos Alexandris wrote:
Agustin:
Nikos, I'm probably missing a message (I do not see which is
exactly the difference between the spot and the modis example),
Hi Agus. The difference is:
Using SPOT (range up to max. 255):
i.pca == prcomp(x, center=TRUE, scale=FALSE)
Using MODIS
Agustin:
Nikos, I'm probably missing a message (I do not see which is
exactly the difference between the spot and the modis example),
Nikos:
Hi Agus. The difference is:
Using SPOT (range up to max. 255):
i.pca == prcomp(x, center=TRUE, scale=FALSE)
Using MODIS (range
On Tue, Mar 31, 2009 at 5:00 PM, Markus Metz
markus.metz.gisw...@googlemail.com wrote:
Nikos Alexandris wrote:
Using MODIS (range varries between bands, up to max. ~5500):
Sure? Valid range: -100 - 16000 for MODIS e.g. Surface Reflectance 8-Day L3
Global 500m.
Remember that you have to
On Tue, 2009-03-31 at 17:13 +0200, Markus Neteler wrote:
On Tue, Mar 31, 2009 at 5:00 PM, Markus Metz
markus.metz.gisw...@googlemail.com wrote:
Nikos Alexandris wrote:
Using MODIS (range varries between bands, up to max. ~5500):
Sure? Valid range: -100 - 16000 for MODIS e.g. Surface
Edzer:
With Markus' help I got access, and corrected/extended quite a few things,
including why R sometimes doesn't print loadings (when they're close to
0).
Great!
I do worry about the wiki that states results are similar between R and
grass, when the differences appear e.g. in the third
Nikos Alexandris wrote:
Nikos:
Hi Agus. The difference is:
Using SPOT (range up to max. 255):
That implied to me that the max possible value of SPOT was used, not the
max observed (often the max observed is also the max possible). In my
experience, when comparing imagery or derived
Nikos:
Hi Agus. The difference is:
Using SPOT (range up to max. 255):
Markus M:
That implied to me that the max possible value of SPOT was used, not the
max observed (often the max observed is also the max possible).
Right. Sorry for not being precise.
In my experience, when
Nikos Alexandris wrote:
The thing is by multiplying by 0.0001 thing are worse concerning the
*eigenvalues* (the eigenvectors are the same):
# using the MODIS bands as they are
r.info -r mod_b2
min=0
max=5504
# use of i.pca gives
r.info -h pca_mod_b267.1
[...]
Data Description:
Nikos:
The thing is by multiplying by 0.0001 thing are worse concerning the
*eigenvalues* (the eigenvectors are the same):
# use of i.pca gives
r.info -h pca_mod_b267.1
[...]
Eigen values, (vectors), and [percent importance]:
PC1 6307563.04 (-0.6353,-0.6485,-0.4192)[98.71%]
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
On Thu, Mar 12, 2009 at 8:26 AM, Hamish hamis...@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)
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.
On Tue, 2009-03-03 at 14:26 +0100, Nikos Alexandris wrote:
On Tue, 2009-03-03 at 13:18 +0100, Markus Neteler wrote:
On Tue, Mar 3, 2009 at 12:26 PM, Hamish hamis...@yahoo.com wrote:
...
it will be really good to finally have all this documented.
I find it hard to follow these long
On Tue, 10 Mar 2009, Nikos Alexandris wrote:
On Tue, 2009-03-03 at 14:26 +0100, Nikos Alexandris wrote:
On Tue, 2009-03-03 at 13:18 +0100, Markus Neteler wrote:
On Tue, Mar 3, 2009 at 12:26 PM, Hamish hamis...@yahoo.com wrote:
...
it will be really good to finally have all this documented.
Nikos:
I've started the PCA grass-wiki page. I am trying to build it... step by
step. Expert advice is always welcome and highly appreciated.
http://grass.osgeo.org/wiki/Principal_Component_Analysis
Roger:
Good, thanks. There you say that you are using some MODIS surface
reflectance
On Tue, Mar 3, 2009 at 12:26 PM, Hamish hamis...@yahoo.com wrote:
...
it will be really good to finally have all this documented.
I find it hard to follow these long mails. Why not enjoying the
GRASS Wiki to stabilize the documentation and comparisons?
Markus
On Tue, 2009-03-03 at 13:18 +0100, Markus Neteler wrote:
On Tue, Mar 3, 2009 at 12:26 PM, Hamish hamis...@yahoo.com wrote:
...
it will be really good to finally have all this documented.
I find it hard to follow these long mails. Why not enjoying the
GRASS Wiki to stabilize the
28 matches
Mail list logo
