<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">Hi Cory,<br>
      <br>
      I'm stuck with this issue :(<br>
      <br>
      1. Changing just the pvti file is not enough: I got a segfault
      using the raw binary files.<br>
      2. I tried also adding a dummy slice at the end of all pieces to
      fix the problem that I reported in one of my previous emails. So,
      instead of changing the PVTI files I tried changing the VTI files.<br>
          2.1. First of all, I had to increase the number of tuples of
      the VtkArray because I was adding  the dummy slice. Since I didn't
      change the whole extent and the piece extent of each VTI file, I
      got this warning message:<font face="Courier New, Courier,
        monospace"><br>
            Warning: In /opt/VTK-6.2.0/Common/DataModel/vtkDataSet.cxx,
        line 421<br>
            vtkImageData (0x3450390): Point array Var with 1 components,
        has 4080501 tuples but there are only 4040100 points</font><br>
          This didn't work either.<br>
          2.2. Afterwards, I tried modifying also the whole extent and
      the piece extent. This seemed to work :)<br>
      <br>
      Despite 2.2 seemed to work, I'm not sure if this is a valid
      solution or just a workaround: If I open a VTI file directly
      (because I want to see the data produced by a MPI process) the
      dummy slice is visualized.<br>
      <br>
      From my point of view, I think that we're trying to address an
      issue of the PVTI format modifying something that works fine (VTI
      files).<br>
      <br>
      Best regards,<br>
      <br>
      Sergi<br>
      <br>
      On 08/12/2015 02:46 PM, Sergi Mateo Bellido wrote:<br>
    </div>
    <blockquote cite="mid:55CB402B.10300@gmail.com" type="cite">
      <meta http-equiv="Context-Type" content="text/html; charset=utf-8">
      <div class="moz-cite-prefix">Cory,<br>
        <br>
        I did not change the VTI files in the first experiment but it
        crashed anyway.<br>
        <br>
        Thanks!<br>
        Sergi<br>
        <br>
        On 08/12/2015 01:55 PM, Cory Quammen wrote:<br>
      </div>
      <blockquote
cite="mid:CAB5Fpx4Z-ZJAigMWtUcE_QnD5AXsO93K07KbFOeAPUk7Hysv5g@mail.gmail.com"
        type="cite">
        <div dir="ltr">Sergi,
          <div><br>
          </div>
          <div>Note that I did *not* change the whole extents of the VTI
            files after I changed the listed extents in the PVTI files.</div>
          <div><br>
          </div>
          <div>Cory</div>
        </div>
        <div class="gmail_extra"><br>
          <div class="gmail_quote">On Wed, Aug 12, 2015 at 1:47 AM,
            Sergi Mateo Bellido <span dir="ltr"><<a
                moz-do-not-send="true"
                href="mailto:sergi.mateo.bellido@gmail.com"
                target="_blank">sergi.mateo.bellido@gmail.com</a>></span>
            wrote:<br>
            <blockquote class="gmail_quote">
              <div>
                <div>Cory,<br>
                  <br>
                  The ascii version is the more stable version and it
                  usually doesn't crash (even the original version, I
                  reported that in one of my first emails). I tried to
                  different things without regenerating the *.vti *.pvti
                  (using the zipped raw version):<br>
                  <br>
                  1. Changing the piece extent in all the *.pvti files
                  to what you proposed in your last email -> segfault<br>
                  2. Changes describe in 1. + changing the whole
                  extension of the VTI files that contain the second
                  piece -> Paraview warns me because the size of the
                  second piece doesn't correspond with the data.<br>
                  <br>
                  vtkXMLImageDataReader (0x419ef70): Cannot read point
                  data array "var" from PointData in piece 0.  The data
                  array in the element may be too short.<br>
                  <br>
                  Following your advice, I will try to generate an extra
                  dummy slice in the first piece. I will let you know
                  the results :)<br>
                  <br>
                  Best,<br>
                  Sergi
                  <div>
                    <div class="h5"><br>
                      <p><br>
                      </p>
                      <br>
                      On 08/11/2015 11:57 PM, Cory Quammen wrote:<br>
                    </div>
                  </div>
                </div>
                <div>
                  <div class="h5">
                    <blockquote type="cite">
                      <div dir="ltr">Actually, you could try one other
                        thing first:
                        <div><br>
                        </div>
                        <div>Change the extents of the pieces in the
                          .pvti files to the following:</div>
                        <div><br>
                        </div>
                        <div>
                          <div>    <Piece Extent="0 200 0 200 0 100"
                            Source="ascii_2_0.vti"/></div>
                          <div>    <Piece Extent="0 200 0 200 100
                            200" Source="ascii_2_1.vti"/></div>
                        </div>
                        <div><br>
                        </div>
                        <div>I'm able to load the files and move between
                          time steps fine doing this.</div>
                        <div><br>
                        </div>
                        <div>Cory</div>
                      </div>
                      <div class="gmail_extra"><br>
                        <div class="gmail_quote">On Tue, Aug 11, 2015 at
                          5:10 PM, Cory Quammen <span dir="ltr"><<a
                              moz-do-not-send="true"
                              href="mailto:cory.quammen@kitware.com"
                              target="_blank">cory.quammen@kitware.com</a>></span>
                          wrote:<br>
                          <blockquote class="gmail_quote">
                            <div dir="ltr">Ideally both pieces would
                              contain a copy of the slice, but I think
                              for the first piece you could just write
                              some dummy values into this extra slice.
                              When you load it, the valid values from
                              the first piece will overwrite the dummy
                              values.
                              <div><br>
                              </div>
                              <div>I'm frankly surprised this extra
                                slice would be needed because the ghost
                                level is set to 0. But I'll admit I
                                don't fully understand this format yet.</div>
                              <span>
                                <div><br>
                                </div>
                                <div>Cory</div>
                              </span></div>
                            <div class="gmail_extra">
                              <div>
                                <div><br>
                                  <div class="gmail_quote">On Tue, Aug
                                    11, 2015 at 3:37 PM, Sergi Mateo
                                    Bellido <span dir="ltr"><<a
                                        moz-do-not-send="true"
                                        href="mailto:sergi.mateo.bellido@gmail.com"
                                        target="_blank">sergi.mateo.bellido@gmail.com</a>></span>
                                    wrote:<br>
                                    <blockquote class="gmail_quote">
                                      <p dir="ltr">Hi Cory,</p>
                                      <p dir="ltr">This extra slice is
                                        the first slice of the second
                                        piece. Should both pieces
                                        contain this slice?</p>
                                      <p dir="ltr">In my case, each MPI
                                        process computes something and
                                        writes its portion of the mesh.
                                        All these portions are
                                        completely disjoint among them.</p>
                                      <p dir="ltr">Thanks!<span><br>
                                          Sergi</span></p>
                                    </blockquote>
                                  </div>
                                  <br>
                                  <br>
                                  <div><br>
                                  </div>
                                </div>
                              </div>
                              <span>-- <br>
                                <div>Cory Quammen<br>
                                  R&D Engineer<br>
                                  Kitware, Inc.</div>
                              </span></div>
                          </blockquote>
                        </div>
                        <br>
                        <br>
                        <div><br>
                        </div>
                        -- <br>
                        <div>Cory Quammen<br>
                          R&D Engineer<br>
                          Kitware, Inc.</div>
                      </div>
                    </blockquote>
                    <br>
                  </div>
                </div>
              </div>
            </blockquote>
          </div>
          <br>
          <br>
          <div><br>
          </div>
          -- <br>
          <div class="gmail_signature">Cory Quammen<br>
            R&D Engineer<br>
            Kitware, Inc.</div>
        </div>
      </blockquote>
      <br>
    </blockquote>
    <br>
  </body>
</html>