[R] Generalized singular value decomposition

[Oana Tomescu]

Thank you for your help!

The GSVD (SVDgen) in PTAk does not perform a generalized singular value 
decomposition for 2 matrices, only for 1. I should have mentioned this - 
sorry.
Are there maybe other packages?

I have also found the last 2 links, but I look for a way to use LAPACK 
with Windows - this issue is unfortunately not addressed in these posts.

Can anyone help me with LAPACK and Windows XP? I have noticed that there 
are already 2 dlls in the R/bin folder (Rblas.dll and Rlapack.dll). But 
I still can not load the libraries. I guess that I still have to 
"install"/do something ...

I would appreciate any help!

Regards,
Oana

Am 30.11.2011 13:18, schrieb Paul Hiemstra:
> Hi,
>
> Googling for "R-help gsvd" leads to a number of interesting entries on
> the mailing list. It seems that GSVD is present in the lapack version
> included with R and can be called (see the mailing list entries).
>
> https://stat.ethz.ch/pipermail/r-help/2004-August/056713.html
> http://r.789695.n4.nabble.com/problem-with-generalized-singular-value-decomposition-using-LAPACK-td834930.html
> http://r.789695.n4.nabble.com/Generalized-SVD-td798791.html
>
> regards,
> Paul
>
> On 11/30/2011 10:51 AM, Oana Tomescu wrote:
>> Hello,
>>
>> I would like to perform a generalized singular value decomposition
>> with R. The only possibility I found is "GSVD" that is based on
>> LAPACK/BLAS. Are there other possibilities too?
>> If not, has anybody used LAPACK/BLAS under Windows XP? How can I
>> install them? Following [1] did not help.
>>
>> I hope this is the right place for my question.
>>
>> Thank you very much!
>> Oana Tomescu
>>
>> [1] http://sites.google.com/site/jivsoft/Home/r-blas-interface
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>
>