Re: [eigen] GivensQR

[ Thread Index | Date Index | More Archives ]

2009/8/18 Benoit Jacob <jacob.benoit.1@xxxxxxxxx>:
> have you benchmarked a
> single vector solve with your class against PartialLU? Because, since
> you're optimizing so much for this kind of use, I wonder if you can
> beat PartialLU; if not, then it's not worth optimizing so much.

Well, then of course, if GivensQR can be made more stable than other
decompositions, especially on singular matrices, that's another reason
to care :)


Mail converted by MHonArc 2.6.19+