On Monday, September 19, 2016 at 9:04:35 AM UTC+2, Ehsan Mihankhah wrote:
> I have been working with Dealii 8.3 for a while and I have been using the
> modified version of step-36 to solve eigenvalue problem with ARPACK. I
> noticed that there is a possibility of working with PARPACK in Dealii 8.4
> and so I upgraded to this version.
> I am trying to get the codes in
> running with MPI and PETSc variables.
So you are combining a installation of 8.4.x with unit tests from the
current master branch.
This is generally a bad idea. I would suggest you configure the current
development version of deal.II from GtiHub, compile and then
try to run those tests by:
ctest -R "arpack"
> I managed to compile the sample codes but the problem is that when I run
> the code I receive errors from ARPACK referring to "info=-9" in Pdnaupd.
> and following "info=-9" I see the error is about initial vector being zero
> and I cannot solve this problem by tuning any parameter it seems.
> Can anybody suggest a solution please?
> Thank you very much,
> Best regards,
> P.S: One important issue here is that the parpack_solver.h that comes with
> the package available on dealii website is not the one on the github and
> therefore it does not recognize some of the functions. So, I
that's not an issue, but the expected behaviour. Deal.II 8.4.x is the 8.4.x
version of the code, whereas "master" branch is the current development
version. The two are not the same by definition.
> copied the version which is available on the github in order to be able to
> get the codes compiled.
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see
You received this message because you are subscribed to the Google Groups
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email
For more options, visit https://groups.google.com/d/optout.