[petsc-users] EPS iteration number

[Xujun Zhao]

The following is another test, which shows the outer iteration is 2, and
MatMult is performed 24 times. So it means the first outer iteration
performs 16, and the subsequent iterations performs 8. (24=16+8)

What parameters can I tune to accelerate the convergence? Will changing ncv
work better?
The following nconv=1 and unconverged value, what do they mean?

2 EPS nconv=1 first unconverged value (error) 0.0140231 (1.32666590e-02)

--->test: EPS solve results:
          the eigenvalue: 0.009301
          Solution method: krylovschur
          Number of iterations of the method: 2
          EPS converged reason : 2

On Wed, Sep 16, 2015 at 2:32 PM, Jose E. Roman <jroman at dsic.upv.es> wrote:

>
> > El 16/9/2015, a las 21:09, Xujun Zhao <xzhao99 at gmail.com> escribió:
> >
> > Hi all,
> >
> > I am solving a max/min eigenvalue problem, and call
> EPSGetIterationNumber() to get the iteration number for convergence.
> However, it always returns 1. This is a little confused, because I can see
> my subroutine of MATOP_MULT operation has been called 6 or 16 times with
> different configurations.
> >
> > The ConvergedReason returns 2, so it looks like the eigenvalue is
> correct.
> > also, it shows
> >  1 EPS nconv=2 first unconverged value (error) 0.0158234 (2.15550644e-03)
> >
> > what does it mean?
> >
> > Thanks.
> >
> > Xujun
>
> In EPS the iteration number refers to “outer” iterations, that is,
> restarts. So in Krylov-Schur with ncv=16 the first outer iteration will
> perform 16 MatMult’s and in subsequent outer iterations it will perform 8
> MatMult’s.
>
> Jose
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20150916/c9cfecd5/attachment.html>