pelahi
commented
3 years ago Hi @swati-gavas I am quite busy for the next few days but will get to this soon. I think the main thing will be to update the hdf5 reader to load the appropriate fields/attributes/data spaces. If I have a spare moment, I'll have a deeper look at the gadget 4 format to implement it. Cheers, Pascal
I am facing some issues when running VELOCIRAPTOR on Gadget4 snapshots. I have elaborated them below.
I have snapshots with 128^3 particles in hdf5 format. On running VELOCIRAPTOR on the hdf5 files I get the following error:
_terminate called after throwing an instance of 'std::invalid_argument' what(): attribute not found NumPart_TotalHighWord
I know VELOCIRAPTOR is not meant to work on GADGET4 hdf5 snapshots (Although it would be great to have this functionality).
I also converted my snapshots to GADGET-2/3 legacy formats and tried running VELOCIRAPTOR again. In case of legacy binary formats code is picking wrong number of particles and it throws the following error.
_ERROR: MPI Process has zero particles associated with it, likely due to too many mpi tasks requested or too coarse a mesh used. Current number of tasks: 4 Current mesh resolution 8^3 Total number of particles loaded 4580513888 Suggested number of particles per mpi processes is > 1e7 Suggested number of mpi processes using 1e7 is 459 Increase mesh resolution or reduce MPI Processes application called MPI_Abort(MPI_COMMWORLD, 8) - process 0
The snapshots, log files and the config file can be accessed through the following link. Note: The folder also contains a README file which describes the file naming convention used.
https://drive.google.com/drive/folders/1gTlBqqyD3SsWorCja0y2MSXS59uuPf3S?usp=sharing
Any help will be highly appreciated Thank you ~Swati