[Paraview] Clean Grid and Average Point Data

[Andy Bauer]

Hi Tristan,

I'm not familiar with deal.II's output format -- could you share a small
sample?

Some other tricks you can try is using the "Pass Point Data" option with
the Point Data to Cell Data filter which will keep your point data around
as well. Then you could use the Pass Arrays filter to get rid of the cell
data that you didn't want generated from the point data before (i.e. the
ones that would overwrite your existing point data when you use the cell
data to point data filter).

As for the programmable filter, you can find more information on it at
http://www.paraview.org/Wiki/Python_Programmable_Filter. There are probably
more examples around the wiki that you can find as well with a little bit
of searching.

Regards,
Andy




On Tue, Jul 29, 2014 at 4:00 AM, Tristan Schlögl <
tristan.schloegl at ltd.uni-erlangen.de> wrote:

> Hi Andy,
>
> thanks for your reply. I tried working with the "point data to cell data"
> and "cell data to point data" filters. Indeed, this is too much averaging.
> It might be a starting point for only the discontinuous field I'm
> interested in, but the filter applies on all data in my model (e.g. I don't
> want the continuous displacement field and others to be averaged). Is there
> a way to specify only a certain dataset the filter is working on?
>
> Regarding the programmable filter, I'm very new to ParaView internals (I'm
> mainly working with the GUI), however I'm a bit experienced with Python.
> Can you recommend some tutorial, where I get a quick introduction in how to
> access the data and modify it?
>
> I believe dealing with this problem in deal.ii directly requires some
> changes in the library itself. Maybe you can also do something with derived
> classes. However, having the full information of a discontinuous field,
> like deal.ii outputs it, is generally not a bad thing to examine results. I
> just sometimes need 'cleaner' pictures and thought that's a post processing
> task, best be handled in ParaView.
>
> Tristan
>
>
> Von: Andy Bauer <andy.bauer at kitware.com>
> Datum: Montag, 28. Juli 2014 16:38
> An: Tristan Schloegl <tristan.schloegl at ltd.uni-erlangen.de>
> Cc: "Paraview at paraview.org" <Paraview at paraview.org>
> Betreff: Re: [Paraview] Clean Grid and Average Point Data
>
> Hi Tristan,
>
> There are a couple of things you can try. The best but most complex is
> probably the programmable filter where you merge the points yourself and do
> the averaging for the point values. The other thing you can try is using
> the point data to cell data filter, then use the clean to grid filter and
> finally use the cell data to point data filter. This may due more averaging
> than you would like for your results. Other than that, did you get any
> feedback on the deal.II mailing list?
>
> Regards,
> Andy
>
>
> On Mon, Jul 28, 2014 at 8:37 AM, Tristan Schlögl <
> tristan.schloegl at ltd.uni-erlangen.de> wrote:
>
>> Hi,
>>
>> I'm using ParaView to visualise finite element simulation results created
>> by the deal.ii C++ library. Deal.ii outputs hex elements such that each
>> element has its own vertices. To illustrate that point a simple cube
>> consisting of 8 elements and 27 nodes has 8*8=64 data points. When I
>> display vector valued data as glyphs I have multiple vectors at the same
>> location resulting from the contributions of neighbouring cells, see
>> http://blogs.fau.de/schloegl/files/2014/07/orig_grid.png .
>>
>> I now want to display only a single vector at each point, namely the
>> average value of all contributions. I tried and applied a "Clean to Grid"
>> filter, see
>> http://blogs.fau.de/schloegl/files/2014/07/clean_grid.png .
>> This is quite good already however, it seems that ParaView only takes one
>> of the vectors for each point and not the average (see top mid vector of
>> the front (z-max) face).
>>
>> Is there any other way to realise what I need in ParaView?
>>
>> Regards
>> Tristan
>>
>>
>> --
>> Dipl.-Ing. Tristan Schlögl
>> Chair of Applied Dynamics
>> University of Erlangen-Nuremberg
>> Haberstrasse 1
>> D-91058 Erlangen, Germany
>> room 01.013
>> phone +49 (0)9131 85 61003
>> fax +49 (0)9131 85 61011
>> url www.ltd.tf.uni-erlangen.de
>>
>> _______________________________________________
>> Powered by www.kitware.com
>>
>> 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
>>
>> Follow this link to subscribe/unsubscribe:
>> http://public.kitware.com/mailman/listinfo/paraview
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20140729/5e7d257e/attachment.html>