Re: New function

[Jason Stover]

On Thu, Mar 26, 2009 at 04:34:52PM +0900, John Darrington wrote:
> On Wed, Mar 25, 2009 at 12:50:54PM -0300, Michel Boaventura wrote:
> 
>      I'm starting to work with an algorith who uses a dataset and a
>      function called singular value decomposition (SVD) to try to find
>      groups on the dataset.
>      So far I'm using the Ben's site to export the .sav files to text, and
>      using Octave to make the SVD. Looking at the GSL doc, they had the SVD
>      implemented
>      in a very efficient way, so I start to think about put this
>      functionality on PSPP. But because PSPP uses SPSS syntax, I'm
>      wondering how new functions like that
>      could be implemented.
> 
> So far as I recall, FACTOR ANALYSIS is the Spss command that deals
> with singular value decomposition.  SVD requires the covariance matrix
> of a data set. Jason is currently working on covariance matrix
> generation.  I'm not sure how complete it is, but it's probably at a
> stage where it can be used for non-categorical variables.

It was until I made a bunch of changes to incorporate categorical
variables and interactions. I'm fixing that now.

-Jason