[Paraview-developers] Source info "memory" problem

[Reuter, Michael A.]

Hi Burlen,

Thanks for having a look at this. The MDEWRebinningPresenter uses the input dataset to gather some metadata and does not modify it. A new VTK dataset is created for the output side of the rebinner plugin based on the supplied parameters. Based on your suggestion, I took a wider look at the input side usage, but the plugin doesn't modify the input dataset in any way. Any other thoughts?

Thanks,
M

From: Burlen Loring <burlen.loring at gmail.com<mailto:burlen.loring at gmail.com>>
Date: Friday, November 22, 2013 11:59 AM
To: Michael Reuter <reuterma at ornl.gov<mailto:reuterma at ornl.gov>>, "paraview-developers at paraview.org<mailto:paraview-developers at paraview.org>" <paraview-developers at paraview.org<mailto:paraview-developers at paraview.org>>
Subject: Re: [Paraview-developers] Source info "memory" problem

Hi Michael,

That sounds like what would happen if your filter delete's its input when it's destroyed. It could be some other modification made to the input too, like removing all the cells or something like that. The rule is don't modify the input data, if need to make a copy. I couldn't tell if you're doing this or not from because you pass the input dataset into MDEWRebinningPresenter constructor. But that'd be the first possibility to eliminate. Are you sure you're not modifying the input dataset?

Burlen


On 11/21/2013 11:29 AM, Reuter, Michael A. wrote:

Hi,

I have a weird problem of a source losing its "memory" after applying a plugin. The source is from a reader that can handle data with or without time step information (see Before_Plugin.jpg showing data with time steps). I then apply a plugin (rebinner) that allows me to alter the number and/or range of the time steps. In this case, I'm leaving both the range and number the same, but changing these doesn't seem to matter to the problem. I wanted to go back to the original source, so I deleted the rebinner plugin, and I found that the original source had lost not only its time info, but the array it was looking at was also gone (see After_Plugin.jpg) and data no longer appears in the view. If I do the same exercise using the reader and rebinner but for data without time steps, the original source doesn't lose its "memory". Therefore, this has something to do with time info handling. I'm having this issue in 3.98.1, but I've tested against 4.1.0-RC1 and see the same issue. I th

 ought abo
ut trying to catch the time step issue in RequestData, but UPDATE_TIME_STEP is returning values, however, they are the wrong ones. The slightly bigger concern is the loss of the signal array from the source. I tried using RequestUpdateExtent to replace the time info in the reader, but this caused more issues and didn't fix the original problem. Is there anything obvious I'm overlooking that I should be investigating further?

The source for the reader is here:
https://github.com/mantidproject/mantid/blob/master/Code/Mantid/Vates/ParaviewPlugins/ParaViewReaders/MDEWNexusReader/vtkMDEWNexusReader.cxx

The source for the rebinner is here:
https://github.com/mantidproject/mantid/blob/master/Code/Mantid/Vates/ParaviewPlugins/ParaViewFilters/MDEWRebinningCutterOperator/vtkMDEWRebinningCutter.cxx

Thanks,
M

Dr. Michael Reuter
Data Analysis and Visualization Group
Neutron Data Analysis and Visualization Division
Oak Ridge National Laboratory

Office: 1-865-241-7216
Fax: 1-865-574-6080
Email: reuterma at ornl.gov<mailto:reuterma at ornl.gov>



_______________________________________________
Paraview-developers mailing list
Paraview-developers at paraview.org<mailto:Paraview-developers at paraview.org>http://public.kitware.com/mailman/listinfo/paraview-developers