ParaView issueshttps://gitlab.kitware.com/paraview/paraview/-/issues2021-03-24T07:58:32-04:00https://gitlab.kitware.com/paraview/paraview/-/issues/20047amrex reader: pick point gaussian representation by default2021-03-24T07:58:32-04:00Utkarsh Ayachitamrex reader: pick point gaussian representation by defaultref: paraview/paraview#19776ref: paraview/paraview#19776https://gitlab.kitware.com/paraview/paraview/-/issues/20046amrex reader: auto detect name of particles sub-directory2021-03-24T07:58:45-04:00Utkarsh Ayachitamrex reader: auto detect name of particles sub-directoryref: paraview/paraview#19776ref: paraview/paraview#19776https://gitlab.kitware.com/paraview/paraview/-/issues/20045File path is not updating on macOS Paraview 5.8.02020-10-27T21:23:31-04:00lorak41File path is not updating on macOS Paraview 5.8.0File path is not updating in the file explorer when using Navigate Up.
(macOS 10.15.5, Paraview 5.8.0)
![Jun-30-2020_17-49-47](/uploads/6c105860a1ad38d5350b4dfb0f78b368/Jun-30-2020_17-49-47.gif)File path is not updating in the file explorer when using Navigate Up.
(macOS 10.15.5, Paraview 5.8.0)
![Jun-30-2020_17-49-47](/uploads/6c105860a1ad38d5350b4dfb0f78b368/Jun-30-2020_17-49-47.gif)5.9 (Fall 2020)Cory Quammencory.quammen@kitware.comCory Quammencory.quammen@kitware.comhttps://gitlab.kitware.com/paraview/paraview/-/issues/20044update loguru2020-06-30T19:21:12-04:00Utkarsh Ayachitupdate loguruv2.1.0 has several useful features like verbosity level for loguru internal messages, ability rename main thread before init, etc. that we can use.v2.1.0 has several useful features like verbosity level for loguru internal messages, ability rename main thread before init, etc. that we can use.5.9 (Fall 2020)Utkarsh AyachitUtkarsh Ayachithttps://gitlab.kitware.com/paraview/paraview/-/issues/20043Dependency in ParaView to libmath.a from embree32020-06-30T11:22:55-04:00Alexander NeumannDependency in ParaView to libmath.a from embree3So in VCPKG CI throws the following error:
`ninja: error: '/mnt/vcpkg-ci/installed/x64-linux/lib/libmath.a', needed by 'bin/pvserver', missing and no known rule to make it`
because somehow there is a hidden dependency on embree3 which ...So in VCPKG CI throws the following error:
`ninja: error: '/mnt/vcpkg-ci/installed/x64-linux/lib/libmath.a', needed by 'bin/pvserver', missing and no known rule to make it`
because somehow there is a hidden dependency on embree3 which install libmath.a. (I mean we could simply add embree3 to ParaViews Build-Depends which solves the issue but that would probably be a hack)
Does somebody here know where this dependency might be picked up by ParaView?https://gitlab.kitware.com/paraview/paraview/-/issues/20042netCDF-CF Reader does not read all variables anymore ...2021-04-29T20:54:32-04:00Niklas RoebernetCDF-CF Reader does not read all variables anymore ...It seems that the ParaView (Master) netCDF-CF reader does not show, load or visualizes all variables anymore. It appears that only the variable that is first in the alphabet is shown as available and can be loaded and visualized. The uni...It seems that the ParaView (Master) netCDF-CF reader does not show, load or visualizes all variables anymore. It appears that only the variable that is first in the alphabet is shown as available and can be loaded and visualized. The units of the other -- not loaded -- variables are shown properly though.
How to test with ParaView Master (5.8.1):
- Load a regular or curvilinear netCDF file using the netCDF-CF reader and click on applyhttps://gitlab.kitware.com/paraview/paraview/-/issues/20041GhostCellsGenerator - Different behaviour in serial and parallel?2020-06-29T21:40:19-04:00Carsten UphoffGhostCellsGenerator - Different behaviour in serial and parallel?Hi,
I have a few simple **parallel** VTU files with cell-wise data, which looks as following:
![cell](/uploads/a83941cd9c67b5a0adb457d340d9fbca/cell.png)
The red and blue triangles belong to different pieces, respectively. After apply...Hi,
I have a few simple **parallel** VTU files with cell-wise data, which looks as following:
![cell](/uploads/a83941cd9c67b5a0adb457d340d9fbca/cell.png)
The red and blue triangles belong to different pieces, respectively. After applying the CellDataToPointData filter I expect to see the following:
![point](/uploads/7ce2509ec3eac2d49104758449255ccf/point.png)
Of course I need ghost cells for this to work, because the red and blue piece are not "connected" in the pvtu file. (Checking whether ghost cells work correctly is actually my intent.) Therefore, I use the GhostCellsGenerator filter, i.e. the sequence of filters
- test.pvtu
- GhostCellsGenerator
- CellDataToPointData
Doing "mpiexec -n 4 pvserver", connecting to the server, and applying the filter sequence gives the desired result. However, doing the same steps without connecting to a server, i.e. using the serial version does not give me the correct result (that is, the blue and red triangle are "disconnected").
Is this behaviour expected or am I doing something wrong? I find this highly confusing.
Note that I'm using the Paraview 5.8 (binary release) and checking the "Piece invariant" option in the CellDataToPointData filter does not change anything.
Thanks!
Carsten
Input files:
[test.pvtu](/uploads/f6619ea7c14873df9cfbbec216f090b0/test.pvtu)
[test1.vtu](/uploads/df9e5689b0cf57e0c95f2ce7ab4011e9/test1.vtu)
[test2.vtu](/uploads/38286f999a1cb327f560a3baf01fa515/test2.vtu)https://gitlab.kitware.com/paraview/paraview/-/issues/20040Some minimum improvement for interactive selection2020-12-02T14:07:22-05:00Richard WitasseSome minimum improvement for interactive selectionOne change would benefit the usability of interactive selection in Paraview: Maintain any chosen Select (cells/points on, …) button from the top toolbar on the view frame. Any selection I have to make now systematically requires me to cl...One change would benefit the usability of interactive selection in Paraview: Maintain any chosen Select (cells/points on, …) button from the top toolbar on the view frame. Any selection I have to make now systematically requires me to click on that Select button over and over again until the selection is OK. I d rather have it always on and consecutively draw a rectangle of clicking cells with the appropriate shift or control addition depending whether I want to add or subtract. Once finished I would end the selection mode by hitting the Select button again (and deactivate the edition of the selection). Anyone would support that idea?5.11 (Spring 2022)https://gitlab.kitware.com/paraview/paraview/-/issues/20039Find Data does not work with strings2021-03-24T07:58:59-04:00Richard WitasseFind Data does not work with stringsI would also wish the criteria would also accept 'string. For instance our FE kernel generate vtkMultiblockDataSet. For each cell we create Cell Data like ElementType, MaterialName, GroupName … which all are strings. Enabling creating se...I would also wish the criteria would also accept 'string. For instance our FE kernel generate vtkMultiblockDataSet. For each cell we create Cell Data like ElementType, MaterialName, GroupName … which all are strings. Enabling creating selection based on those names will be extremely powerful imo for begineers and medium level users.https://gitlab.kitware.com/paraview/paraview/-/issues/20038Build failure with external HDF5 (v1.12) and PV 5.8 build2020-07-09T16:17:34-04:00Alexander NeumannBuild failure with external HDF5 (v1.12) and PV 5.8 buildObserved error in VCPKG CI:
```
D:\buildtrees\paraview\src\4fb6e6800f-2efd43d7fc\Utilities\VisItBridge\databases\readers\M3DC1\avtM3DC1FileFormat.C(2057): error C2660: 'H5Oget_info_by_name3': function does not take 4 arguments
D:\instal...Observed error in VCPKG CI:
```
D:\buildtrees\paraview\src\4fb6e6800f-2efd43d7fc\Utilities\VisItBridge\databases\readers\M3DC1\avtM3DC1FileFormat.C(2057): error C2660: 'H5Oget_info_by_name3': function does not take 4 arguments
D:\installed\x64-windows-static\include\H5Opublic.h(188): note: see declaration of 'H5Oget_info_by_name3'
D:\buildtrees\paraview\src\4fb6e6800f-2efd43d7fc\Utilities\VisItBridge\databases\readers\M3DC1\avtM3DC1FileFormat.C(2116): error C2660: 'H5Oget_info3': function does not take 2 arguments
D:\installed\x64-windows-static\include\H5Opublic.h(187): note: see declaration of 'H5Oget_info3'
```
Either Paraview needs to set the required HDF5 preprocessor definitions to select the proper HDF5 api or VisItBridge needs updating for the new API.https://gitlab.kitware.com/paraview/paraview/-/issues/20037Catalyst: support pvsm state files as pipelines2021-03-24T07:59:19-04:00Utkarsh AyachitCatalyst: support pvsm state files as pipelinesWith support for extract generators, a ParaView state includes relevant information for generation of extracts in situ. Typically, the current workflow is save out Python state file and load that in situ. We should support non-python sta...With support for extract generators, a ParaView state includes relevant information for generation of extracts in situ. Typically, the current workflow is save out Python state file and load that in situ. We should support non-python state files (i.e. pvsm) as well. This will make it possible for simulations to not depend on Python -- which has been a pain point for some HPC use-cases.https://gitlab.kitware.com/paraview/paraview/-/issues/20036Extract Generators: support loading pvsm state file in batch to generate extr...2021-03-24T07:59:27-04:00Utkarsh AyachitExtract Generators: support loading pvsm state file in batch to generate extractsWe already support saving python state file and then running it in pvbatch/pvpython to generate extracts. Can't we support the same of pvsm? pvsm states are generally more robust and hence provide a nicer alternative for folks who merely...We already support saving python state file and then running it in pvbatch/pvpython to generate extracts. Can't we support the same of pvsm? pvsm states are generally more robust and hence provide a nicer alternative for folks who merely want to generate extracts without doing any additional fine tuning to the state.https://gitlab.kitware.com/paraview/paraview/-/issues/20035Display changes after applying rendering, all windows re-scaled2020-06-27T14:40:30-04:00Roger RogerDisplay changes after applying rendering, all windows re-scaledI am running Paraview 5.8 on a Windows 10 system with an NVIDIA GTX1650 GPU. The program loads fine but after uploading a geometry, for example, and then selecting "Apply" to render it, everything in the Paraview environment gets re-sca...I am running Paraview 5.8 on a Windows 10 system with an NVIDIA GTX1650 GPU. The program loads fine but after uploading a geometry, for example, and then selecting "Apply" to render it, everything in the Paraview environment gets re-scaled to various sizes or magnifications. Thus, I am unable to select any menu options as they don't appear visually anymore. Some selections are in the field of view but you don't know that you are selecting them because the button is not visible.[paraview-windows10-display-issue.pdf](/uploads/a1ec98aa15c772c384c2c475cc49b576/paraview-windows10-display-issue.pdf)https://gitlab.kitware.com/paraview/paraview/-/issues/20034Nvidia Index building issue2021-06-22T19:48:02-04:00Lucas GasparinoNvidia Index building issueGreetings,
When trying to build the master version of ParaView with pvNvidiaIndex option, both the configure and compilation steps run OK. After that, the ctest fails for the plugin, and trying to use paraview with it results in the fol...Greetings,
When trying to build the master version of ParaView with pvNvidiaIndex option, both the configure and compilation steps run OK. After that, the ctest fails for the plugin, and trying to use paraview with it results in the following error:
nvindex: PVPLN init error: libnvindex.so: cannot open shared object file: No such file or directory
I did set "/path/to/paraview/lib" in my LD_LIBRARY_PATH. I had no warnings or error during compilation or "ccmake ..". Am I missing an option? I just turned to option ON, which works for all other plugins. I can't find this library anywhere, and I'm assuming it must have been built by the paraview build process. What I do find is the ../lib/paraview-5.8/plugins/pvNvidiaIndex/pvNVIDIAIndeX.so and libvtkIndeXRepresentations.so. I've tried various options in the ccmake, with the same results.
Thank you for your attention!https://gitlab.kitware.com/paraview/paraview/-/issues/20033Building with OptiX2021-04-20T17:04:22-04:00Lucas GasparinoBuilding with OptiXGreetings!
It's my first time building paraview with raytracing, and I'd like to use nvidia optix for that, as it is suggested as an alternative to ospray. How would i go about configuring this in cmake? I do have Optix installed, and a...Greetings!
It's my first time building paraview with raytracing, and I'd like to use nvidia optix for that, as it is suggested as an alternative to ospray. How would i go about configuring this in cmake? I do have Optix installed, and a Quadro P4200 gpu running cuda 11.0 and latest driver. Sorry if this is a silly question, but i can't find a proper answer anywhere.
In any case, thank you very much!https://gitlab.kitware.com/paraview/paraview/-/issues/20032Extract Generators: Show progress dialog when generating extracts2020-10-29T14:39:18-04:00Utkarsh AyachitExtract Generators: Show progress dialog when generating extractsOn hitting Save Extracts, let's show the same progress (and cancel) dialog that we show when saving animations.On hitting Save Extracts, let's show the same progress (and cancel) dialog that we show when saving animations.5.9 (Fall 2020)Utkarsh AyachitUtkarsh Ayachithttps://gitlab.kitware.com/paraview/paraview/-/issues/20031Show spack gitlab-ci results on Cdash2020-08-31T14:57:56-04:00Utkarsh AyachitShow spack gitlab-ci results on CdashCurrently, we have a gitlab-ci build for spack. However, the results from that are not shown in CDash which makes it hard to catch failures. We need to show spack build status (at the very least) on CDash.
One solution is as follows:
* ...Currently, we have a gitlab-ci build for spack. However, the results from that are not shown in CDash which makes it hard to catch failures. We need to show spack build status (at the very least) on CDash.
One solution is as follows:
* Under `Utilities/` add a new dir which is new CMake project for building ParaView using Spack i.e. it should merely have custom commands that runs spack using params specified to build. (see `.gitlab-ci.yml` for spack commands to use to build).
* This project can also have tests to confirm build was successful
* add CTestConfig.cmake for this project to be same as `ParaView/CTestConfig.cmake` so ctest will drop results on the same site as ParaView.
Now, one should be able to use ctest to build using spack and summit results.
Finally, update gitlab-ci.yaml to use ctest to build/test/submit instead of directly using spack.Christos TsolakisChristos Tsolakishttps://gitlab.kitware.com/paraview/paraview/-/issues/20030Calculator Filter Malfunctions for Dataset Names starting with a Number2021-09-21T21:43:55-04:00Joel A. KuleszaCalculator Filter Malfunctions for Dataset Names starting with a NumberThe Calculator filter malfunctions with dataset names that begin with a number.
For example, entering a calculation that should return a Result based on a dataset named `test_1` behaves correctly but a dataset named `1_test` throws an...The Calculator filter malfunctions with dataset names that begin with a number.
For example, entering a calculation that should return a Result based on a dataset named `test_1` behaves correctly but a dataset named `1_test` throws an error (see below). This behavior has been confirmed with legacy VTK (see attached) and HDF5-based XDMF files. The presence/absence of the underscore in the name has no effect.
Other filters appear to operate correctly with either dataset (e.g., Threshold).
Steps to reproduce on macOS 10.14.6:
1. Launch ParaView 5.8.0
2. File -> Open -> `block_4x4.vtk` -> Apply
3. Filters -> Calculator
4. Enter calculation formula `test_1`
5. Apply (works)
6. Change calculation formula to `1_test`
7. Apply (fails)
CC: @ethan.stam @patchett2002
![Screen_Shot_2020-06-24_at_14.24.21](/uploads/70e944b50814cce6aba12c609a2d39a6/Screen_Shot_2020-06-24_at_14.24.21.png)
[block_4x4.vtk](/uploads/6a096396a4b549d2810d82cb663b24ea/block_4x4.vtk)Spiros TsalikisSpiros Tsalikishttps://gitlab.kitware.com/paraview/paraview/-/issues/20029Multiple Expressions in Find Data forgets half of the query2020-09-07T13:18:38-04:00Jean M. FavreMultiple Expressions in Find Data forgets half of the queryI wish to extract points from an unstructured grid. So, I just applied a CleantoGrid to Wavelet to make a reproducer test. The picture says it all:
Wavelet+CleantoGrid
my query should be "(RTData <= 100) & (points[:,2] > 0)" but ParaView...I wish to extract points from an unstructured grid. So, I just applied a CleantoGrid to Wavelet to make a reproducer test. The picture says it all:
Wavelet+CleantoGrid
my query should be "(RTData <= 100) & (points[:,2] > 0)" but ParaView drops the first part of it. This is v5.8.1-RC1
![FailedSelection](/uploads/24cec0462dd2c263761d4d2cd6d352c5/FailedSelection.png)5.9 (Fall 2020)https://gitlab.kitware.com/paraview/paraview/-/issues/20028Text size mismatch2021-04-07T11:08:02-04:00Utkarsh AyachitText size mismatchWe want to get to point where all text in ParaView renders consistently.
**Text A:**
* Sources | Text, Change text to RTData, Apply
**Text B:**
* sources | Wavelet, Apply, Color by RTData, move scalar bar so that it's horizontal i.e. R...We want to get to point where all text in ParaView renders consistently.
**Text A:**
* Sources | Text, Change text to RTData, Apply
**Text B:**
* sources | Wavelet, Apply, Color by RTData, move scalar bar so that it's horizontal i.e. RTData oriented the same was as Text A.
* For Color Map Editor panel, edit color legend params by clicking ![image](/uploads/6551b81776cd727b4bef1593f71329e6/image.png) and change Title Font Parameters to match the Text A. i.e Arial, size = 18.
The two texts don't look the same.
![image](/uploads/eb1a99e7e85f1beef1daebaa3a8f36dc/image.png)