simulation.hdf5 (https://www.tng-project.org/data/docs/specifications/#sec4a) is really convenient (and amazing!). I am really curious how this can be done.
However, I found that some items in this file are all zero. For example:
with h5py.File('/home/tnguser/sims.TNG/TNG100-1/simulation.hdf5', 'r') as fp:
assert (fp['Groups/99/Group/Group_M_Crit200'][()] == 0).all()
assert (fp['Groups/99/Group/Group_M_Mean200'][()] == 0).all()
assert (fp['Groups/99/Group/GroupFirstSub'][()] == 0).all()
and the same for TNG100-1-Dark/simulation.hdf5. Is it expected or not? If expected, then it might be nice to have some explanation about the limitation in the documentation.
I must apologize! Just read the document again and find the possible cause. My h5py's version ('2.7.1', /opt/conda/lib/python3.6/site-packages/h5py) is too low. I will update it first then try again.
I will keep this post for reference in case anyone made the same mistake as me.