[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: octave application to experimental spectra
From: |
Juan Pablo Carbajal |
Subject: |
Re: octave application to experimental spectra |
Date: |
Mon, 26 Mar 2012 20:59:20 +0200 |
On Mon, Mar 26, 2012 at 8:48 PM, Bryan Lepore <address@hidden> wrote:
> I need to know if octave will do the following, before I work out exactly how:
>
> I am interested in the singular value decomposition of a large number
> of experimental recordings of spectra - essentially x/y data, where y
> changes but x is the same for each y.
>
> Will octave do this if I figure out how to [1] load one spectrum per
> row in a matrix, and then [2] run svd as in the manual?
>
> -Bryan
> _______________________________________________
> Help-octave mailing list
> address@hidden
> https://mailman.cae.wisc.edu/listinfo/help-octave
Bryan,
Octave can do SVD in large sets (how large depends more on your
machine than on Octave). Additionally, you can get fastICA2.5 and also
do ICA analysis (similar to SVD/PCA). Furthermore, the devel folder of
linear-algebra package from Octave Forge has non-negative matrix
factorization, which is like PCA, but the components are restricted to
be positive.
For more academic examples of Octave uses, check this out
http://octave.org/wiki/index.php?title=Publications_using_Octave
I hope to you see one of your publications there soon!
:D
--
M. Sc. Juan Pablo Carbajal
-----
PhD Student
University of Zürich
http://ailab.ifi.uzh.ch/carbajal/