Closed pkufourier closed 1 month ago
I cannot reproduce what you are seeing on my Linux machine with Paraview. Previously, you mentioned you are using Windows. Is it possible that's the issue? Note that long
on Windows is 4 bytes. I wonder if this is a Paraview issue on Windows.
Sorry for the late reply. Yes, you are correct, it's the problem for the Windows-version of Paraview. I have also tested this code with the Linux Paraview, and the data loading works well even for 1024 cubic data. I will report this issue for Paraview. And I also find a similar question on the Paraview forum (https://discourse.paraview.org/t/paraview-large-dataset-visualisation/15499). Maybe it is the same reason.
test_readwrite.zip Please see the above testing code. AMReX is compiled with MPI+OMP, CPU only, double precision. The problem is, when the cubic box have a grid number n_cell=420 and I load the generated plt file with Paraview 5.13.0 (the latest version), statistics in Paraview shows the variable v5 is not resolved correctly as following:
If I set n_cell=400, the result for v5 seems to be normal:
I have tested the cases with AMReX 24.01 and 24.09. The results are the same. So I'm not sure whether the problem is inside WriteMultiLevelPlotfile or Paraview.