<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.2900.2802" name=GENERATOR></HEAD>
<BODY style="MARGIN: 4px 4px 1px; FONT: 10pt Tahoma">
<DIV>Mathieu,</DIV>
<DIV> </DIV>
<DIV>OK. I did a simple test program that brings in an vtkXMLUnstructured Grid and does a surface filter and then renders and quits. I run this test code on data that has been meshed with tets and another that has been meshed with polyhedrons. I emphasize, this is the same data that has been meshed two different ways. When I run the tetrahedron data through my program it executes in 4 seconds, when I execute the polyhedron mesh it takes 4 minutes and 26 seconds. I ran them both on callgrind and in the polyhedron case it is spending a lot of time in LUFactorLinearSystem and LUSolveLinearSystem.</DIV>
<DIV> </DIV>
<DIV>I put the code, data, and callgrind output on my ftp site.</DIV>
<DIV> </DIV>
<DIV><A href="https://lars.netl.doe.gov/public/dotson/">https://lars.netl.doe.gov/public/dotson/</A></DIV>
<DIV> </DIV>
<DIV>Let me know what you think.</DIV>
<DIV> </DIV>
<DIV>Thanks,</DIV>
<DIV> </DIV>
<DIV>Brian Dotson</DIV>
<DIV>National Energy Technology Laboratory<BR><BR>>>> "Mathieu Malaterre" <mathieu.malaterre@kitware.com> 4/13/2006 5:26 PM >>><BR></DIV>
<DIV style="COLOR: #000000">Hi Brian,<BR><BR> Let's try to do things in order. Can you exhibit a case where the use <BR>of vtkConvexPointCell is really slow in a small VTK/C++ code. Ideally <BR>you should be able to run valgrind+callgrind to analyze what can/should <BR>be optimized.<BR> A much more complex solution would be to use the Generic framework, I <BR>don't know if you are familiar with it, but you can find it in <BR>VTK/GenericFiltering. Unfortunately it is not as well integrated in <BR>ParaView. But this is one of our goal.<BR><BR>Anyway let me know how it goes, and what is your time frame ?<BR>Thanks<BR>Mathieu<BR><BR>Brian Dotson wrote:<BR>> Hi all,<BR>> <BR>> <BR>> <BR>> I am working with a group of CFD researchers that are using ParaView to <BR>> view their data. Last year we investigated how ParaView would handle <BR>> polyhedral meshes. I took their data and created a VTK file that <BR>> consisted of an unstructured grid with convex point sets for cells. <BR>> This worked, but when the cell counts and complexities got higher the <BR>> performance dropped considerably.<BR>> <BR>> <BR>> <BR>> I started looking at this problem again, and was wondering if a new VTK <BR>> cell type would help this problem.<BR>> <BR>> <BR>> <BR>> My reasoning is that when I am creating the convex point set cells, I am <BR>> throwing a lot of useful information away. When I receive the <BR>> information from the CFD simulation I know the following:<BR>> <BR>> <BR>> <BR>> 1. The number of points in each face of the cell.<BR>> <BR>> 2. What points make up each face of the cell.<BR>> <BR>> 3. The points that make up the face are in a specific order.<BR>> <BR>> 4. All of the points are on the surface, no internal points.<BR>> <BR>> 5. The cell number on each side of the face.<BR>> <BR>> 6. The number of faces per cell.<BR>> <BR>> <BR>> <BR>> I was thinking a cell that consisted of a set of faces might be more <BR>> efficient in our application. Each face would consist of a list of <BR>> point in a specified order where the normal vector would point away from <BR>> the cell. <BR>> <BR>> <BR>> <BR>> I am willing to do the work on this, but does this make sense?<BR>> <BR>> <BR>> <BR>> Please give me your thoughts.<BR>> <BR>> <BR>> <BR>> Thanks,<BR>> <BR>> <BR>> <BR>> Brian Dotson<BR>> <BR>> National Energy Technology Laboratory<BR>> <BR>> <BR>> <BR>> <BR>> ------------------------------------------------------------------------<BR>> <BR>> _______________________________________________<BR>> vtk-developers mailing list<BR>> vtk-developers@vtk.org<BR>> <A href="http://www.vtk.org/mailman/listinfo/vtk-developers">http://www.vtk.org/mailman/listinfo/vtk-developers</A><BR><BR>_______________________________________________<BR>vtk-developers mailing list<BR>vtk-developers@vtk.org<BR><A href="http://www.vtk.org/mailman/listinfo/vtk-developers">http://www.vtk.org/mailman/listinfo/vtk-developers</A><BR><BR></DIV></BODY></HTML>