[petsc-dev] DMPlex overlap redistribution

Matthew Knepley knepley at gmail.com
Fri Mar 8 11:50:56 CST 2024


On Thu, Mar 7, 2024 at 12:08 AM Adrian Croucher <a.croucher at auckland.ac.nz>
wrote:

> hi Matt,
>
> With a bit more experimenting, I find I get the same sort of error in
> DMPlexStratify() if I run a very simple program that just reads in a
> DMPlex from file and then stratifies it (no dual porosity, no
> distribution etc.), i.e.:
>
>    call DMPlexCreateFromFile(PETSC_COMM_WORLD, 'gminc_1d.exo', 'mesh', &
>         PETSC_TRUE, dm, ierr)
>    call DMPlexStratify(dm, ierr)
>
> The error again looks like this:
>
> [0]PETSC ERROR: Petsc has generated inconsistent data
> [0]PETSC ERROR: New depth 0 range [10,54) overlaps with depth 0 range
> [10,54)
>
> Would you expect that to work?
>

Here is what is happening in this case. The DMPlexCreateFromFile() call
already stratifies the mesh.
Then the call DMPlexStratify() attempts to do this again. There is a check
in DMPlexCreateDepthStratum()
that checks for overlap. If we want to call this again, we first need to
wipe out the depth label. I wasn't sure I wanted DMPlexStratify() to do
this automatically, but maybe it should.

  Thanks,

     Matt


> - Adrian
>
> --
> Dr Adrian Croucher
> Senior Research Fellow
> Department of Engineering Science
> Waipapa Taumata Rau / University of Auckland, New Zealand
> email: a.croucher at auckland.ac.nz
> tel: +64 (0)9 923 4611
>
>

-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener

https://urldefense.us/v3/__https://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!egE85226em0pMIxZzhtCDoypqqpVkgjeDShU6xwNDaUaSg_Yalo45qmkBBW2k1X4NfXTdLKw0pfSGPXryDoa$  <https://urldefense.us/v3/__http://www.cse.buffalo.edu/*knepley/__;fg!!G_uCfscf7eWS!egE85226em0pMIxZzhtCDoypqqpVkgjeDShU6xwNDaUaSg_Yalo45qmkBBW2k1X4NfXTdLKw0pfSGAOd7xGB$ >
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20240308/ae58e4b9/attachment.html>


More information about the petsc-dev mailing list