ParaView issueshttps://gitlab.kitware.com/paraview/paraview/-/issues2024-03-07T20:47:49-05:00https://gitlab.kitware.com/paraview/paraview/-/issues/18228ospray mode of GL volume mapper has noticeable process boundaries2024-03-07T20:47:49-05:00David E. DeMarleospray mode of GL volume mapper has noticeable process boundariesConnect to a parallel server, wavelet, representation to volume, volume rendering mode to OSPRay Based.
The process boundaries will become visible.Connect to a parallel server, wavelet, representation to volume, volume rendering mode to OSPRay Based.
The process boundaries will become visible.5.13 (Summer 2024)David E. DeMarleDavid E. DeMarlehttps://gitlab.kitware.com/paraview/paraview/-/issues/18227image downsampling during client server interaction wonky with ospray2024-03-07T20:47:49-05:00David E. DeMarleimage downsampling during client server interaction wonky with osprayIn client server mode, the default is to downsample images in client server mode when the user interacts. We don't tell ospray to compensate for that though. The result is that you see a very magnified corner of the image while you are m...In client server mode, the default is to downsample images in client server mode when the user interacts. We don't tell ospray to compensate for that though. The result is that you see a very magnified corner of the image while you are moving the camera around.5.13 (Summer 2024)David E. DeMarleDavid E. DeMarlehttps://gitlab.kitware.com/paraview/paraview/-/issues/18212Save data removes partial arrays2021-04-28T09:53:20-04:00W. Alan ScottSave data removes partial arraysWhen we are writing data with partial arrays, we should not just add 0's to the variables. The partial arrays should remain partial. Here is how to replicate:
* Linux, built-in server, 5.5.0
* Read in can.exo, block 1, all vars on, ex...When we are writing data with partial arrays, we should not just add 0's to the variables. The partial arrays should remain partial. Here is how to replicate:
* Linux, built-in server, 5.5.0
* Read in can.exo, block 1, all vars on, except ACCL and EQPS. Saved as canA.exo.
* Read in can.exo, block 2, all vars on. Saved as canB.exo.
* Read in canA.exo and canB.exo. All vars on. Highlighted both. Group Datasets. I now have partial arrays.
* Saved canWithPartialArrays.exo.
When I read this file, there are no partial arrays! The arrays from canA have been filled in with zeros.https://gitlab.kitware.com/paraview/paraview/-/issues/18209QtTesting Check RenderView screenshot freeze paraview2021-03-23T18:41:19-04:00Mathieu Westphal (Kitware)QtTesting Check RenderView screenshot freeze paraviewUsing the QtTesting check capability on the render view, wich trigger a screenshot, seems to freeze ParaView.
(tested with nvidia)
How to reproduce :
* Open ParaView
* Tools -> Record Test
* Click on "Check" button
* Click on Rende...Using the QtTesting check capability on the render view, wich trigger a screenshot, seems to freeze ParaView.
(tested with nvidia)
How to reproduce :
* Open ParaView
* Tools -> Record Test
* Click on "Check" button
* Click on Render View
* Save a screenshot
* ParaView freeze
```
0x00007fffb0fa4786 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
(gdb) bt
#0 0x00007fffb0fa4786 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#1 0x00007fffb0fbf5e8 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#2 0x00007fffb0fbb21d in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#3 0x00007fffb0fb7c25 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#4 0x00007fffb0f87ba8 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#5 0x00007fffb0f89b03 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#6 0x00007fffb0d4f053 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#7 0x00007fffb0d225c4 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#8 0x00007fffb0d231dd in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#9 0x00007fffb0c35a39 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#10 0x00007fffb0f068da in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#11 0x00007fffb0dce22b in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#12 0x00007fffb0dcbb57 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#13 0x00007fffb0dddd68 in ?? () from /usr/lib/libnvidia-glcore.so.390.48
#14 0x00007fffef96e10e in vtkOpenGLQuadHelper::Render() ()
from /home/glow/work/paraview/paraview_build/lib/libvtkRenderingOpenGL2-pv5.5.so.1
#15 0x00007fffef90866a in vtkOpenGLFXAAFilter::ApplyFilter() ()
from /home/glow/work/paraview/paraview_build/lib/libvtkRenderingOpenGL2-pv5.5.so.1
#16 0x00007fffef9088b1 in vtkOpenGLFXAAFilter::Execute(vtkOpenGLRenderer*) ()
from /home/glow/work/paraview/paraview_build/lib/libvtkRenderingOpenGL2-pv5.5.so.1
#17 0x00007fffef984702 in vtkOpenGLRenderer::UpdateGeometry() ()
from /home/glow/work/paraview/paraview_build/lib/libvtkRenderingOpenGL2-pv5.5.so.1
#18 0x00007fffef980430 in vtkOpenGLRenderer::DeviceRender() ()
from /home/glow/work/paraview/paraview_build/lib/libvtkRenderingOpenGL2-pv5.5.so.1
#19 0x00007fffe89675bf in vtkRenderer::Render() ()
```https://gitlab.kitware.com/paraview/paraview/-/issues/18208QtTesting Check overlay not working with new QVTKOpenGLWidget2021-03-23T18:41:37-04:00Mathieu Westphal (Kitware)QtTesting Check overlay not working with new QVTKOpenGLWidgetThe new implementation of QVTKOpenGLWidget from !2117 prevent QtTesting to correctly show the check overlay on the RenderView, as described [here] (https://blog.kitware.com/new-check-feature-in-qttesting-available-in-paraview/).
How to ...The new implementation of QVTKOpenGLWidget from !2117 prevent QtTesting to correctly show the check overlay on the RenderView, as described [here] (https://blog.kitware.com/new-check-feature-in-qttesting-available-in-paraview/).
How to reproduce :
* Open ParaView
* Tools -> Record Test
* Click on "Check" button
* Hover on Render View
* nothing happened
The check overlay on the old implementation does work has it can be tested on the Opacity transfer function widget for example.https://gitlab.kitware.com/paraview/paraview/-/issues/18203Exodus file read failures message has become confusing2024-03-07T20:47:49-05:00W. Alan ScottExodus file read failures message has become confusingIf a user has a bad read, we have moved from a clean, clear error message to trash. Here is how to replicate:
* Linux, 5.5.0, remote server (possibly replicates built-in server)
* Have someone else copy a multiple file Exodus dataset so...If a user has a bad read, we have moved from a clean, clear error message to trash. Here is how to replicate:
* Linux, 5.5.0, remote server (possibly replicates built-in server)
* Have someone else copy a multiple file Exodus dataset somewhere. Have them open permissions on the directory, and the zero'th file. Have them lock you out of the rest of the files. I used file g1s1. Utkarsh and/or Cory should have this one. (If not, ask me for it).
* File/ open/ open this dataset. Apply. It will fail as follows:
Exodus Library Warning/Error: [vtkexodusII_ex_check_valid_file_id]
ERROR: In "ex_inquire_internal", the file id -1 was not obtained via a call to "ex_open" or "ex_create".
It does not refer to a valid open exodus file.
Aborting to avoid file corruption or data loss or other potential problems.
* Linux, 5.4.1, remote server (possibly replicates built-in server)
* File/ open/ open this dataset. It will fail as follows:
ERROR: In .../ParaView5/VTK/IO/Exodus/vtkExodusIIReader.cxx, line 3874
vtkExodusIIReaderPrivate (0x16e0200): Unable to open .../g1s1-deleteMe/g1s1-10fps.e.16.10 for reading
ERROR: In .../ParaView/5.4.1-sb/ParaView5/VTK/IO/Exodus/vtkExodusIIReader.cxx, line 5456
vtkExodusIIReader (0x16b1070): Unable to open file .../g1s1-deleteMe/g1s1-10fps.e.16.10 to read metadata5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/18200VCR controls stay disabled even after disconnecting from a live Catalyst session2024-03-07T20:47:49-05:00Utkarsh AyachitVCR controls stay disabled even after disconnecting from a live Catalyst sessioncc: @acbauer @demarlecc: @acbauer @demarle5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/18196Texture is split when volume redering with multicore2021-03-23T18:37:22-04:00Gwang-Jin LeeTexture is split when volume redering with multicoreHello.
I'm using ParaView 5.4.1 version with multicore. But I met a problem.
When I use pvpython with single core `pvpython PlotScala.py`, I got the expected results.
![pvpython](/uploads/f4e706a6aa6c2f710bf5cfc16378637f/pvpytho...Hello.
I'm using ParaView 5.4.1 version with multicore. But I met a problem.
When I use pvpython with single core `pvpython PlotScala.py`, I got the expected results.
![pvpython](/uploads/f4e706a6aa6c2f710bf5cfc16378637f/pvpython.png)
But when I use pvpatch with muticore `mpiexec -np 32 pvbatch PlotScala.py`, I found the textures of wavefront obj were split:
![pvbatch](/uploads/f885508a432a771e023126aad7bde565/pvbatch.png)
And I found that this problem is related to the volume rendering. When I didn't use volume rendering (like surface), there was no problem with single and multi-core.
I attached my all related files.
Thank you.
[dust.vtk](/uploads/b879136137e2d524e075ca9934719ee0/dust.vtk)
[PlotScala.py](/uploads/5a51546af26a4398cf20f95d4f710b06/PlotScala.py)
[Buildings.png](/uploads/8141d3eabfa1a59bbbd73dd6aab16821/Buildings.png)
[Buildings.obj](/uploads/7552d692a70d706fa06008ed958fd5b4/Buildings.obj)https://gitlab.kitware.com/paraview/paraview/-/issues/18195pvd doesn't support vth2021-03-23T18:32:02-04:00Burlen Loringpvd doesn't support vthOPening a PVD file pointing to vth datasets results in an the following error. It seems to be an oversight in the list of supported file extensions, see vtkXMLCollectionReaderInternals::ReaderList in vtkXMLCollectionReader.cxx. I suspect...OPening a PVD file pointing to vth datasets results in an the following error. It seems to be an oversight in the list of supported file extensions, see vtkXMLCollectionReaderInternals::ReaderList in vtkXMLCollectionReader.cxx. I suspect it would be an easy fix to add it.
```
ERROR: In /home/sensei/sc17/software/paraview/builds/ParaView/ParaViewCore/VTKExtensions/Default/vtkXMLCollectionReader.cxx, line 414
vtkPVDReader (0x4d12530): Could not determine the data type for the first dataset. Please make sure this file format is supported.
ERROR: In /home/sensei/sc17/software/paraview/builds/ParaView/VTK/Common/ExecutionModel/vtkExecutive.cxx, line 784
vtkPVCompositeDataPipeline (0x4eb7c60): Algorithm vtkPVDReader(0x4d12530) returned failure for request: vtkInformation (0x4ec7600)
Debug: Off
Modified Time: 165109
Reference Count: 1
Registered Events: (none)
Request: REQUEST_DATA_OBJECT
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
```
[pvd_bug.tar.gz](/uploads/7f677de70676e45b13f797fceff374e7/pvd_bug.tar.gz)https://gitlab.kitware.com/paraview/paraview/-/issues/18194LS-Dyna reader should use ParaView's open-directory feature2022-06-09T14:09:54-04:00David ThompsonLS-Dyna reader should use ParaView's open-directory featureLS-Dyna databases are actually directories, not files. It accepts a "d3plot" file for historical reasons but it would be better to call `SetDatabaseDirectory()` on it rather than `SetFileName()` which currently accepts a file and then tr...LS-Dyna databases are actually directories, not files. It accepts a "d3plot" file for historical reasons but it would be better to call `SetDatabaseDirectory()` on it rather than `SetFileName()` which currently accepts a file and then tries to discern its containing directory.https://gitlab.kitware.com/paraview/paraview/-/issues/18192Point Gaussian Scale Array Magnitude does not work with short and char array2021-05-10T14:31:05-04:00Mathieu Westphal (Kitware)Point Gaussian Scale Array Magnitude does not work with short and char arrayShort and Char Array scaling the Point Gaussian by magnitude are not working correctly
How to reproduce :
- Create a PointSource, nPoints 10, Radius 1, Apply
- Add a RandomAttributes, select any of Bit, Char, Unsigned Char, Short, Unsi...Short and Char Array scaling the Point Gaussian by magnitude are not working correctly
How to reproduce :
- Create a PointSource, nPoints 10, Radius 1, Apply
- Add a RandomAttributes, select any of Bit, Char, Unsigned Char, Short, Unsigned Short
- Color by RandomPointVectors, Magnitude
- Representation -> Point Gaussian
- Check Point Gaussian -> Use Scale Array
- Select RandomPointVectors -> Magnitude ase a scale array
- All point dissapears, unexecpted
- Change Random Attributes type to double
- All point reappears, correctly scaled.https://gitlab.kitware.com/paraview/paraview/-/issues/18185Improve efficiency of CSV reader2021-03-23T18:33:31-04:00Greg VernonImprove efficiency of CSV readerParaView's CSV reader is not particularly efficient for large files. See issue #18180 for discussion.
While CSV is not a particularly good format for large data, they are simple to generate and (for better or worse) are fairly common. ...ParaView's CSV reader is not particularly efficient for large files. See issue #18180 for discussion.
While CSV is not a particularly good format for large data, they are simple to generate and (for better or worse) are fairly common. A couple ideas might be to assume all values are numeric, or to allow the user to specify the format string. An additional thought would be to have a GUI to help the user determine how they'd like the data to be imported, akin to Matlab's file importer (see image below).
![image](/uploads/6c0e02421cc073bd88516a2538cb17fa/image.png)https://gitlab.kitware.com/paraview/paraview/-/issues/18184Key controls in line source are not working2024-03-07T20:47:49-05:00Cory Quammencory.quammen@kitware.comKey controls in line source are not working* Open ParaView 5.5.0
* Add **Line Source**
* Change view mode to 3D
* Zoom in on data.
* Click on an endpoint and hold down the x key. The selected point should only move along the x axis, but it starts behaving oddly. Same for y and z....* Open ParaView 5.5.0
* Add **Line Source**
* Change view mode to 3D
* Zoom in on data.
* Click on an endpoint and hold down the x key. The selected point should only move along the x axis, but it starts behaving oddly. Same for y and z.
This works fine in ParaView 5.4.15.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/18178OSPRay Materials need cleanup2024-03-07T20:47:49-05:00W. Alan ScottOSPRay Materials need cleanupA few observations on the picture at the bottom of #18016. I am slowly learning how to visualize materials with OSPRay. So, you have a learning curve occurring here also. Wood looks spectacular. If you haven't created texture coordin...A few observations on the picture at the bottom of #18016. I am slowly learning how to visualize materials with OSPRay. So, you have a learning curve occurring here also. Wood looks spectacular. If you haven't created texture coordinates, picture looks bad, but there is no warning. We need a warning if no texture coordinates. Further, I am pretty sure we will soon need to come up with other ways to calculate texture coordinates.
I actually like the Aluminum much better now, as I used a better object. disk_out_ref.exo has too large of cells. Also, the textureMapToCylinder uses cell faces, rather than points. This leaves big flat areas. Fixed by using Sources/Cylinder.
I see lots of little random pixels. This was corrected by changing Samples per Pixel. 7 works pretty well. We need to do two things:
- Figure out where these random pixels are coming from. I believe they are an error. We shouldn't mask them - we should fix them.
- Have ParaView move to 1 Sample per Pixel when mouse is down, Change Samples per Pixel to say 7 or 9 when mouse is up.
How I tested:
- Sphere. Change resolutions to 64X64.
- Transform (move the sphere to the side)
- TextureMapToSphere. Wood.
- Cylinder. Change Resolution to 64x64. (Cylinder appears to already have texture coordinates.)
- Transform. Rotate the Cylinder 90 degrees.5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/18177Plot Data over Time fails in parallel2024-03-07T20:47:48-05:00W. Alan ScottPlot Data over Time fails in parallelThe Plot Data over Time filter does not present aggregate data in parallel, but rather gives results seperately for each rank. This needs to be fixed, to present information from a DATASET, not a RANK!
* 5.5.0, Linux, remote server (I ...The Plot Data over Time filter does not present aggregate data in parallel, but rather gives results seperately for each rank. This needs to be fixed, to present information from a DATASET, not a RANK!
* 5.5.0, Linux, remote server (I will use 16 ranks, but this replicates with 2.)
* Open g1s1-10fps.e.16.[0-15]. All variables on. Apply.
* Plot Data over Time. Apply.
Notice that you are given the option to turn on ranks 0 through 15. This should not be presented as ranks, but rather the whole dataset!5.13 (Summer 2024)https://gitlab.kitware.com/paraview/paraview/-/issues/18173Bad clipping range for far placed objects2021-03-23T18:35:03-04:00RustemBad clipping range for far placed objectsHere is a state file showing the effects of bad clipping range when there are at least 2 far placed objects. Uses wavelet. It is creating a box, a wavelet, then translates wavelet to position (6e+6,6e+6,0).
- Linux, 5.5 from site.
- Lo...Here is a state file showing the effects of bad clipping range when there are at least 2 far placed objects. Uses wavelet. It is creating a box, a wavelet, then translates wavelet to position (6e+6,6e+6,0).
- Linux, 5.5 from site.
- Load state file: broken-clipping-range.pvsm.
Nothing will appear.
- +Z and zoom to data on created Box.
Now the box is visible.
- Left mouse button, drag left.
You now see the cube flickering.
The same happens every time when there are at least 2 far placed object (any vtk object type is affected). And it is also present when rendering in VTK.[broken-clipping-range.pvsm](/uploads/f82e6e95a906fcd8b9f1eba13575afb7/broken-clipping-range.pvsm)https://gitlab.kitware.com/paraview/paraview/-/issues/18164Comparative View Inspector doesn't get traced in Python2021-03-23T18:35:31-04:00Utkarsh AyachitComparative View Inspector doesn't get traced in Pythonchanges made to the Comparative View inspector panel don't get traced in Python.changes made to the Comparative View inspector panel don't get traced in Python.https://gitlab.kitware.com/paraview/paraview/-/issues/18158paraview.simple.Box() fails to draw box at right location with MPI2021-03-23T18:35:42-04:00ufdupparaview.simple.Box() fails to draw box at right location with MPIDear all,
paraview.simple.Box(), when run in parallel, fails to draw boxes at the correct coordinates as specified by boxLength and boxCenter. In serial mode, however, the box is drawn at the correct location. At the moment, a ad-hoc fi...Dear all,
paraview.simple.Box(), when run in parallel, fails to draw boxes at the correct coordinates as specified by boxLength and boxCenter. In serial mode, however, the box is drawn at the correct location. At the moment, a ad-hoc fix is to draw the box by using 'paraview.simple.Line()' multiple times. This seems to work both in parallel and serial.
For debugging purposes, I attach a script (example.py) that reproduces the problem.
I am running ParaView-5.4.1 on Debian 9. I am using mpiexec and pvbatch that come together with ParaView.
My shell command for running the script is: /opt/ParaView-5.4.1/bin/mpiexec -np 4 pvbatch example.py
In 'example.py' I basically create two boxes with same center but one has twice the length of the other.
If you run this code with 1 processor you should see that the output (example.png) displays two concentric boxes. Increasing the processor count (anything >1) seems to cause the center of one of the boxes to be misplaced.
This problem only appears when the boxCenter is something different than [0.,0.,0.]. Modifying line 8 of the attached script and setting boxCenter to [0.,0.,0.] results in
correct behavior.![example](/uploads/020217e462a225c40878835bb88dbe05/example.png)
[example.py](/uploads/758d0d4fd26c0c5eca23f8ba85830420/example.py)
Kind Regards,
Tiagohttps://gitlab.kitware.com/paraview/paraview/-/issues/18139PolyLine source remove point feature not working as expected2021-03-23T18:36:22-04:00Mathieu Westphal (Kitware)PolyLine source remove point feature not working as expectedPoints removal does not works correctly with a PolyLine source
How to reproduce :
- Open ParaView
- Add a PolyLineSource, apply
- Ctrl Click on the line to add a point, drag it a bit to break the line, apply, works as expected
- Sh...Points removal does not works correctly with a PolyLine source
How to reproduce :
- Open ParaView
- Add a PolyLineSource, apply
- Ctrl Click on the line to add a point, drag it a bit to break the line, apply, works as expected
- Shift click on the same point, point is removed from the widget, but not from the point list
- Cannot Apply, no modification has be done on the property.https://gitlab.kitware.com/paraview/paraview/-/issues/18138Clip/Slice with a surface2023-11-17T09:55:19-05:00Mathieu Westphal (Kitware)Clip/Slice with a surfaceIt is currently not possible to clip/slice with a vtkPolyData surface dataset in ParaView.
It is possible to do it with a Python Programmable Filter, but this kind of script writing or out of reach for most users.
Here is a Programmabl...It is currently not possible to clip/slice with a vtkPolyData surface dataset in ParaView.
It is possible to do it with a Python Programmable Filter, but this kind of script writing or out of reach for most users.
Here is a Programmable Filter script that do a clip with a surface (pv 5.10). Do not forget to set output type to `vtkUnstructuredGrid` before the first apply.
```plaintext
import paraview.modules.vtkPVVTKExtensionsFiltersGeneral as ext
inpMain = self.GetInputDataObject( 0, 0 )
inp = self.GetInputDataObject( 0, 1 )
out = self.GetOutput()
clipper = ext.vtkPVMetaClipDataSet()
clipper.SetInputDataObject(inpMain)
clipper.SetInsideOut(True)
function = vtk.vtkImplicitPolyDataDistance()
function.SetInput(inp)
clipper.SetDataSetClipFunction(function)
clipper.Update()
outData = clipper.GetOutputDataObject(0)
out.DeepCopy(outData)
```
Here is a Programmable Filter script that do a slice with a surface (pv 5.10). Do not forget to set output type to `vtkPolyData` before the first apply.
```plaintext
import paraview.modules.vtkPVVTKExtensionsFiltersGeneral as ext
inpMain = self.GetInputDataObject( 0, 0 )
inp = self.GetInputDataObject( 0, 1 )
out = self.GetOutput()
slicer = ext.vtkPVMetaSliceDataSet()
slicer.SetInputDataObject(inpMain)
function = vtk.vtkImplicitPolyDataDistance()
function.SetInput(inp)
slicer.SetDataSetCutFunction(function)
slicer.Update()
outData = slicer.GetOutputDataObject(0)
out.DeepCopy(outData)
```
Implementing something similar in C++ shouldn't be too hard.