[Paraview] Adding Filters to Catalyst

Thu Jun 7 10:25:11 EDT 2018

Hi Gregg,

Catalyst will use all of the MPI ranks in the global communicator unless
you pass it another MPI communicator upon initialization. It sounds like
the data isn't getting constructed in the adaptor the way you think it is.
You may want to try the Process Id Scalars filter to see where the data is
really located and add some print statements in the adaptor to verify the
behavior.

Best,
Andy

On Wed, Jun 6, 2018 at 2:08 PM Gregg Streuber <
Gregg.Streuber at aero.bombardier.com> wrote:

> Hey Andy,
>
>
>
> Apologies, I must have hit “reply” without thinking, rather than reply
> all.  I’ve actually managed to resolve that issue, it turned out to be a
> configuration issue with my build of OpenGL.
>
>
>
> I do have another question though.  I’ve been playing around processing
> some data using Catalyst’s python scripting in parallel and the results
> aren’t quite what I expected.  What I’ve been using is a toy 50,000,000
> node cube mesh with an arbitrary velocity field, and having Catalyst do
> some assorted processing – taking cuts etc.  I’ve divided the mesh into 256
> blocks using the VTKMultiBlockDataSet structure, on 256 cores.  But when I
> load the results into Paraview it only indicates 12 VTKBlocks on the
> VTBBlockColors filter. There are 256 divisions in the mesh, but 244 of them
> are duplicates, coloured with one of the 1st 12 colours.
>
>
>
> Adding new lines to the color map, mapped to values >11, doesn’t appear to
> change anything.
>
>
>
> I’m also not seeing much in the way  of performance variation when I use
> 13-256 cores, it’s only when I drop below 12 that things take appreciably
> longer.
>
>
>
> Is Paraview by default only taking advantage of 12 cores at a time?  Is
> this something I can change?
>
>
>
> *From:* Andy Bauer [mailto:andy.bauer at kitware.com]
> *Sent:* Wednesday, June 06, 2018 10:18 AM
> *To:* Gregg Streuber <Gregg.Streuber at aero.bombardier.com>; Paraview (
> paraview at paraview.org) <paraview at paraview.org>
> *Subject:* Re: [Paraview] Adding Filters to Catalyst
>
>
>
> Hi Gregg,
>
>
>
> Please keep the conversations on the mailing list so that everyone can
> participate (also, emails get lost in my inbox so that can also delay a
> quick response).
>
>
>
> I believe that the Catalyst examples should compile properly with all of
> the Catalyst editions (though I haven't tested them). If you were able to
> use an OpenGL with some editions then that should work for any edition that
> depends on OpenGL. Are you working with a Catalyst edition build directory
> or an install directory? If you give me some more information on how to
> reproduce the error I can try and track things down.
>
>
>
> Also, what version of ParaView are you using?
>
>
>
> Cheers,
>
> Andy
>
>
>
> On Fri, Jun 1, 2018 at 3:37 PM Gregg Streuber <
> Gregg.Streuber at aero.bombardier.com> wrote:
>
> Hi Andy,
>
>
>
> Thanks for the information.  I had a full compile of Paraview running on
> the side just in case, I was just curious if there was an easy and
> efficient way to customize a Catalyst install.
>
>
>
> I’ve since gotten Paraview compiled, but when I try to test the
> functionality using some of the included Catalyst examples
> (Examples/Catalyst) they don’t work.  Specifically several of them throw
> “undefined reference” errors during compilation, referring to a couple of
> OpenGL files in the Paraview install.  These examples worked fine when I
> compiled them with the Catalyst build I was testing before.  Do these
> examples require a specific build of OpenGL (or are they incompatible with
> certain other modules)?
>
>
>
> Thanks,
> Gregg
>
>
>
> *From:* Andy Bauer [mailto:andy.bauer at kitware.com]
> *Sent:* Friday, June 01, 2018 11:46 AM
> *To:* Gregg Streuber <Gregg.Streuber at aero.bombardier.com>
> *Cc:* paraview at public.kitware.com
> *Subject:* Re: [Paraview] Adding Filters to Catalyst
>
>
>
> Hi Gregg,
>
>
>
> You can always use a full build of ParaView for Catalyst to make sure you
> get all of the filters available during Catalyst runs.
>
>
>
> There is information at
> https://blog.kitware.com/paraview-catalyst-editions-what-are-they/ which
> gives details about the editions but there's currently not a lot of details
> on how to make your own edition. Editing the cmake.sh script to add in more
> VTK filters will likely not work as the Catalyst editions don't even have
> all of the source code from the ParaView source.
>
>
>
> As for finding the full list of filters and modules available, the blog
> post has some of that but it's changed some. My recommendation would be to
> build an edition with Python enabled and then use pvpython to see what's
> available. Alternatively you could look at files like filters.xml in the
> Catalyst source code to see what's there. It's not ideal but allows you to
> check on stuff before actually running a simulation.
>
>
>
> Best,
>
> Andy
>
>
>
> On Fri, Jun 1, 2018 at 10:10 AM Gregg Streuber <
> Gregg.Streuber at aero.bombardier.com> wrote:
>
> Hi All,
>
>
>
> I’ve recently installed the most full-featured version of Catalyst
> available on the website, but I noticed it’s lacking a lot of filters that
> I need very much (streamlines, for example).  I understand the installed
> packages are fully customizable, but can’t find any good documentation on
> how to go about doing this.   Is this as simple as enabling different
> options in the cmake.sh script? And if so, is there a complete list of the
> filters in paraview and the modules that contain them somewhere?
>
>
>
> Cheers,
>
> Gregg
>
> _______________________________________________
> Powered by www.kitware.com
>
> ParaView discussion is moving! Please visit
> https://discourse.paraview.org/ for future posts.
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ParaView Wiki at:
> http://paraview.org/Wiki/ParaView
>
> Search the list archives at: http://markmail.org/search/?q=ParaView
>
> Follow this link to subscribe/unsubscribe:
> https://public.kitware.com/mailman/listinfo/paraview
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://public.kitware.com/pipermail/paraview/attachments/20180607/4a1a6f98/attachment.html>