|
|
Ublas : |
From: Thomas Lemaire (thomas.lemaire_at_[hidden])
Date: 2006-05-08 16:27:58
Thanks for all your answers !
Indeed, I thought the gelsd lapack routine was also returning U,V and singular
values from the SVD, but it does not... And I need this to compute the
covariance matrix of the least-squares minimized vector... So I will rather
use gesdd routine, for which bindings already exist. Any advice or experience
on this topic are welcomed !
Also, I would suggest to add a link to
http://cvs.sourceforge.net/viewcvs.py/*checkout*/boost-sandbox/boost-sandbox/libs/numeric/bindings/doc/index.html
somewhere in the wiki, this doc is very usefull :-)
thomas
On Monday 08 May 2006 16:36, Thomas Lemaire wrote:
> Dear list,
>
> I am planning to add ublas bindings to lapack gelsd routine (Linear Least
> Squares (LLS) Problems using divide-and-conquer SVD). Googling for this, I
> found the post
> http://groups.yahoo.com/group/ublas-dev/message/1289
>
> It was posted in late 2003 and I cannot find anything in the current
> boost-sandbox CVS. Does any body have more information about this ?
>
> Is there a short guide on howto add such a binding ? Looking at the already
> wrapped routines may be enough...
>
> How can I contribute to the bindings library ? by posting the code on this
> list ? It would be easier for me if my code could end up in the official
> bindings CVS...
>
> cheers,
-- thomas http://www.laas.fr/~tlemaire