TemporalParticlesToPathlines fails with AMReX particle data: "The input dataset did not have a valid DATA_TIME_STEPS information key"
Attached is a .zip file with data for a single particle following a helical trajectory. particle_pathlines_data.zip
- particle_pathlines_vtp.pvsm uses the data in the particle_vtp directory, which is in .vtp format.
- particle_pathlines_amrex.pvsm uses the data in the particle_amrex directory, which is in AMReX format. (This contains some other field data, most of which is zero.)
- When using .vtp format, the “Temporal Particles To Pathlines” filter works as expected.
- When using AMReX format, “Temporal Particles To Pathlines” fails with the following error message:
( 138.181s) [paraview ]vtkTemporalPathLineFilt:343 ERR| vtkTemporalPathLineFilter (0x1a07a500): The input dataset did not have a valid DATA_TIME_STEPS information key
( 138.182s) [paraview ] vtkExecutive.cxx:753 ERR| vtkPVCompositeDataPipeline (0x1a08c7e0): Algorithm vtkTemporalPathLineFilter(0x1a07a500) returned failure for request: vtkInformation (0x1f5f6d90)
Debug: Off
Modified Time: 2442784
Reference Count: 1
Registered Events: (none)
Request: REQUEST_DATA
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
FROM_OUTPUT_PORT: 0
( 138.182s) [paraview ]vtkTemporalPathLineFilt:343 ERR| vtkTemporalPathLineFilter (0x1a07a500): The input dataset did not have a valid DATA_TIME_STEPS information key
( 138.182s) [paraview ] vtkExecutive.cxx:753 ERR| vtkPVCompositeDataPipeline (0x1a08c7e0): Algorithm vtkTemporalPathLineFilter(0x1a07a500) returned failure for request: vtkInformation (0x1f5f6d90)
Debug: Off
Modified Time: 2442931
Reference Count: 1
Registered Events: (none)
Request: REQUEST_DATA
ALGORITHM_AFTER_FORWARD: 1
FORWARD_DIRECTION: 0
FROM_OUTPUT_PORT: 0
Additional information:
- Alexander D. Kaiser ran into a potentially related issue in May 2020. His solution was to export from Visit to vtu format and create a .pvd file. I did not have to create a .pvd file in order to get a working result with data in .vtp format.
- Link to the ParaView discourse thread where I first posted about this.