<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>I read everything again, I think I did not understand you at
      first. The first solution is to modify the DAG, so that the
      rightmost cell is linked to the leftmost face, right ? To do that,
      do I have to manually edit the DAG (the mesh is read from a file)
      ? If so, the mesh connectivity is like the one of a torus, then
      how does it work with the cells/faces coordinates ?</p>
    <p>Now I think the second method may be more straightforward. What's
      the idea ? Get the mapping with DMGetLocalToGlobalMapping, then
      create the mapping corresponding to the periodicity with
      ISLocalToGlobalMappingCreate, and finally
      ISLocalToGlobalMappingConcatenate ? I'm not sure this is the way,
      and I did not find something like DMSetLocalToGlobalMapping to
      restore the modified mapping.</p>
    <p><br>
    </p>
    <p>Pierre<br>
    </p>
    <br>
    <div class="moz-cite-prefix">On 15/10/21 15:33, Pierre Seize wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:8ff9d951-6958-aa1a-b875-e7488bb6b30b@onera.fr">
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      <p>When I first tried to handle the periodicity, I found the <tt>DMPlexCreateBoxMesh</tt>
        function (I cannot find the cylinder one).</p>
      <p>From reading the sources, I understand that we do some work
        either in <tt>DMPlexCreateCubeMesh_Internal</tt> or with <tt>DMSetPeriodicity</tt>.</p>
      <p>I tried to use <tt>DMSetPeriodicity</tt> before, for example
        with a 2x2 box on length 10. I did something like:<br>
      </p>
      <tt> const PetscReal maxCell[] = {2, 2};</tt><tt><br>
      </tt><tt>const PetscReal L[] = {10, 10};</tt><tt><br>
      </tt><tt>const DMBoundaryType bd[] = {DM_BOUNDARY_PERIODIC,
        DM_BOUNDARY_PERIODIC}</tt><tt>;<br>
      </tt><tt>DMSetPeriodicity(dm, PETSC_TRUE, maxCell, L, bd);<br>
      </tt><tt>// or:<br>
      </tt><tt>DMSetPeriodicity(dm, PETSC_TRUE, NULL, L, bd);</tt><br>
      <p>but it did not work: <br>
      </p>
      <tt>VecSet(X, 1);<br>
        DMGetLocalVector(dm, &locX);</tt><tt><br>
      </tt><tt>VecZeroEntries(locX);</tt><tt><br>
      </tt><tt>DMGlobalToLocalBegin(dm, X, INSERT_VALUES, locX);</tt><tt><br>
      </tt><tt>DMGlobalToLocalEnd(dm, X, INSERT_VALUES, locX);</tt><br>
      <tt>VecView(locX, PETSC_VIEWER_STDOUT_WORLD);</tt>
      <p>but the ghost cells values are all 0, only the real cells are
        1. So I guess <tt>DMSetPeriodicity</tt> alone is not sufficient
        to handle the periodicity. Is there a way to do what I want ?
        That is set up my DMPlex in a way that  <tt>DMGlobalToLocalBegin/</tt><tt>DMGlobalToLocalEnd</tt>
        do exchange values between procs AND exchange the periodic
        values?</p>
      <p><br>
      </p>
      <p>Thanks for the help</p>
      <p><br>
      </p>
      <p>Pierre<br>
      </p>
      <br>
      <div class="moz-cite-prefix">On 15/10/21 14:03, Matthew Knepley
        wrote:<br>
      </div>
      <blockquote type="cite"
cite="mid:CAMYG4Gk+ibpcAdCrCByUebYt6p=dRS9fE3xjA04VeMkmQS65wg@mail.gmail.com">
        <div dir="ltr">
          <div dir="ltr">On Fri, Oct 15, 2021 at 7:31 AM Pierre Seize
            <<a href="mailto:pierre.seize@onera.fr"
              moz-do-not-send="true">pierre.seize@onera.fr</a>>
            wrote:<br>
          </div>
          <div class="gmail_quote">
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex">
              <div bgcolor="#FFFFFF">
                <p>It makes sense, thank you. In fact, both ways seems
                  better than my way. The first one looks the most
                  straightforward. Unfortunately I do not know how to
                  implement either of them. Could you please direct me
                  to the corresponding PETSc functions ?<br>
                </p>
              </div>
            </blockquote>
            <div>The first way is implemented for example in
              DMPlexCreateBoxMesh() and DMPlexCreateCylinderMesh(). The
              second is not implemented since</div>
            <div>there did not seem to be a general way to do it. I
              would help if you wanted to try coding it up.</div>
            <div><br>
            </div>
            <div>  Thanks,</div>
            <div><br>
            </div>
            <div>    Matt </div>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px
              0.8ex;border-left:1px solid
              rgb(204,204,204);padding-left:1ex">
              <div bgcolor="#FFFFFF">
                <p>Pierre<br>
                </p>
                <br>
                <div>On 15/10/21 13:25, Matthew Knepley wrote:<br>
                </div>
                <blockquote type="cite">
                  <div dir="ltr">
                    <div dir="ltr">On Fri, Oct 15, 2021 at 7:08 AM
                      Pierre Seize <<a
                        href="mailto:pierre.seize@onera.fr"
                        target="_blank" moz-do-not-send="true">pierre.seize@onera.fr</a>>
                      wrote:<br>
                    </div>
                    <div class="gmail_quote">
                      <blockquote class="gmail_quote" style="margin:0px
                        0px 0px 0.8ex;border-left:1px solid
                        rgb(204,204,204);padding-left:1ex">Hi,<br>
                        <br>
                        I'm writing a code using PETSc to solve NS
                        equations with FV on an <br>
                        unstructured mesh. Therefore I use DMPlex.<br>
                        <br>
                        Regarding periodicity, I manage to implement it
                        this way:<br>
                        <br>
                           - for each couple of boundaries that is
                        linked with periodicity, I <br>
                        create a buffer vector with an
                        ISLocalToGlobalMapping<br>
                        <br>
                           - then, when I need to fill the ghost cells
                        corresponding to the <br>
                        periodicity, the i "true" cell of the local
                        vector fills the buffer <br>
                        vector on location i with
                        VecSetValuesBlockedLocal, then <br>
                        VecAssemblyBegin/VecAssemblyEnd ensure each
                        value is send to the correct <br>
                        location thanks to the mapping, then the i
                        "ghost" cell of the local <br>
                        vector reads the vector on location i to get
                        it's value.<br>
                        <br>
                        <br>
                        It works, but it seems to me there is a better
                        way, with maybe PetscSF, <br>
                        VecScatter, or something I don't know yet. Does
                        anyone have any advice ?<br>
                      </blockquote>
                      <div><br>
                      </div>
                      <div>There are at least two other ways to handle
                        this. First, the method that is advocated in</div>
                      <div>Plex is to actually make a periodic geometry,
                        meaning connect the cells that are meant</div>
                      <div>to be connected. Then, if you partition with
                        overlap = 1, PetscGlobalToLocal() will fill in</div>
                      <div>these cell values automatically.</div>
                      <div><br>
                      </div>
                      <div>Second, you could use a non-periodic
                        geometry, but alter the LocalToGlobal map such</div>
                      <div>that the cells gets filled in anyway. Many
                        codes use this scheme and it is straightforward</div>
                      <div>with Plex just by augmenting the map it makes
                        automatically.</div>
                      <div><br>
                      </div>
                      <div>Does this make sense?</div>
                      <div><br>
                      </div>
                      <div>  Thanks,</div>
                      <div><br>
                      </div>
                      <div>     Matt</div>
                      <div> </div>
                      <blockquote class="gmail_quote" style="margin:0px
                        0px 0px 0.8ex;border-left:1px solid
                        rgb(204,204,204);padding-left:1ex"> Pierre Seize<br>
                      </blockquote>
                    </div>
                    -- <br>
                    <div dir="ltr">
                      <div dir="ltr">
                        <div>
                          <div dir="ltr">
                            <div>
                              <div dir="ltr">
                                <div>What most experimenters take for
                                  granted before they begin their
                                  experiments is infinitely more
                                  interesting than any results to which
                                  their experiments lead.<br>
                                  -- Norbert Wiener</div>
                                <div><br>
                                </div>
                                <div><a
                                    href="http://www.cse.buffalo.edu/%7Eknepley/"
                                    target="_blank"
                                    moz-do-not-send="true">https://www.cse.buffalo.edu/~knepley/</a><br>
                                </div>
                              </div>
                            </div>
                          </div>
                        </div>
                      </div>
                    </div>
                  </div>
                </blockquote>
                <br>
              </div>
            </blockquote>
          </div>
          <br clear="all">
          <div><br>
          </div>
          -- <br>
          <div dir="ltr" class="gmail_signature">
            <div dir="ltr">
              <div>
                <div dir="ltr">
                  <div>
                    <div dir="ltr">
                      <div>What most experimenters take for granted
                        before they begin their experiments is
                        infinitely more interesting than any results to
                        which their experiments lead.<br>
                        -- Norbert Wiener</div>
                      <div><br>
                      </div>
                      <div><a
                          href="http://www.cse.buffalo.edu/%7Eknepley/"
                          target="_blank" moz-do-not-send="true">https://www.cse.buffalo.edu/~knepley/</a><br>
                      </div>
                    </div>
                  </div>
                </div>
              </div>
            </div>
          </div>
        </div>
      </blockquote>
      <br>
    </blockquote>
    <br>
  </body>
</html>