[petsc-users] reg: Eigenvectors of general complex eigen value problem

[venkatesh g]

Ok. I used -st_type sinvert only,

I am comparing it eigs after normalization. Even now the eigenvectors
differ.

Whether I should use a different eps solver ?

Pls let me know.




On Sun, Mar 23, 2014 at 9:46 PM, Jose E. Roman <jroman at dsic.upv.es> wrote:

>
> El 23/03/2014, a las 12:56, venkatesh g escribió:
>
> > Hi,
> >
> > I am using SLEPC example EX7.C, I am using complex binary matrices A and
> B written from matlab using Petscbinarywrite.m.
> >
> > I ran the example using "./ex7 -f1 A -f2 B -eps_type krylovschur
> -st_type sinvert -evecs out.mat -eps_smallest_magnitude"
>
> You should not use -st_type sinvert together with -eps_smallest_magnitude.
> Read the documentation.
>
> >
> > My plotted my eigenvectors using Petscbinaryread and comparing them to
> MATLAB's eigs command.
> >
> > It is wrong. It is right only if I use real matrices A and B.
> >
> > I compiled Petsc for complex scalars. How to get the correct
> eigenvectors for complex matrices, please note my eigenvalue is right but
> not my eigenvector.
>
> How do you know the eigenvectors are wrong? Did you check the residuals?
> If you want to compare the eigenvectors to the output of eigs, you may need
> to normalize them in the same way, for instance with v=v/max(v).
>
> Jose
>
> >
> > Any help is greatly appreciated! Kindly help me out.
> >
> > cheers,
> > Venkatesh
> > Sr. Research Fellow
> >
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20140324/738ae2ca/attachment.html>