Subject: Re: [ublas] announcement: ublasJama (ublas port of Jama, including SVD and eigenvalue decomposition)
From: David Bellot (david.bellot_at_[hidden])
Date: 2010-05-10 05:43:48
well, even if it's inconsistent, it will serve as a base to the
implementation of those algorithms into uBlas. I do not expect them to
integrate "for free" but thanks to Frederic, we have a good starting point
now. Some work needs to be done so that to make things more generic.
2010/5/10 Frédéric Devernay <frederic.devernay_at_[hidden]>
> Le 9 mai 2010 à 02:12, Nasos Iliopoulos a écrit :
> If this is the case, I think there should be a short discussion of the
> interface uBlas may adopt to communicate with such algorithms. I am not sure
> if the Function objects would be appropriate. Maybe simple function calls to
> algorithms would be more generic (and if somebody wants to wrap it into a
> functor it would be elementary). Anyway the maintainers will decide how they
> feel this should be done.
> I agree with that, I picked up the interface from Jama, but it's clearly
> inconsistent with the rest of ublas. I'll leave this discussion to ublas
> maintainers, though.
> ublas mailing list
> Sent to: david.bellot_at_[hidden]
-- David Bellot, PhD david.bellot_at_[hidden] http://david.bellot.free.fr