Remi Lehe

I agree.

I can do a PR on this.

Agree with @DavidSagan's suggestion. @DavidSagan : Do you want to do a corresponding PR?

@ax3l Actually, if it's okay with you, I was planning to implement this with #161. I can try to do it by the end of today, so that it does...

Here is the way it is currently done in openPMD-viewer: ``` python if filename[-3:] == '.h5' or filename[-5:] == '.hdf5': # Extract the iteration, using regular expressions (regex) regex_match =...

@prkkumar Recently shared with me a case which is unstable even with `nx = 512`: [inputs.txt](https://github.com/ECP-WarpX/WarpX/files/9721023/inputs.txt) 

Unfortunately, it looks like this the bug does not show up in 2D: [inputs.txt](https://github.com/ECP-WarpX/WarpX/files/9721080/inputs.txt) 

Interestingly, it looks like the bug is suppressed (in 3D) when changing ``` amr.blocking_factor_x = 128 64 amr.blocking_factor_y = 128 64 ``` to ``` amr.blocking_factor_x = 128 128 amr.blocking_factor_y =...

I also tried removing PMLs with: ``` do_pml_Lo[lev] = amrex::IntVect::TheZeroVector(); do_pml_Hi[lev] = amrex::IntVect::TheZeroVector(); ``` in `Source/Initialization/WarpXInitData.cpp`, but it still shows the instability.

Interestingly, the instability is also removed when using decomposition which is such that the MR patch is not transversally decomposed:  This was achieved with: ``` amr.max_grid_size_x...