I am trying to locate a halo of about 10**13 solar mass at z=0.7 (so snap 059 for TNG50) and find all the gas/galaxy/DM data within the cubes that contains the halo and has cube length of 2000 ckpc/h. I ran some particle searching codes that have worked on my computer before, which basically searches all the snap_059.x.hdf5 for the gas/galaxy/dm particles within the cube. I did this because we want to see some cosmic web stuff, and I think the gas particles associated with cosmic web may be in the inner or outer fuzz and cannot be searched by subhalo.
The problem is that I am trying to do this on Jupyter Lab so that I don't need to download all the data to my computer for searching, and Juyter Lab doesn't have enough memory to complete the searching. The searching may also take a lot of time if the memory issue is fixed. I wonder if there is some faster method to extract these data?
Dylan Nelson
21 Apr
I think your script will work with a minor modification: load the data in chunks (you never need to load it all at once).
Hi all,
I am trying to locate a halo of about 10**13 solar mass at z=0.7 (so snap 059 for TNG50) and find all the gas/galaxy/DM data within the cubes that contains the halo and has cube length of 2000 ckpc/h. I ran some particle searching codes that have worked on my computer before, which basically searches all the snap_059.x.hdf5 for the gas/galaxy/dm particles within the cube. I did this because we want to see some cosmic web stuff, and I think the gas particles associated with cosmic web may be in the inner or outer fuzz and cannot be searched by subhalo.
The problem is that I am trying to do this on Jupyter Lab so that I don't need to download all the data to my computer for searching, and Juyter Lab doesn't have enough memory to complete the searching. The searching may also take a lot of time if the memory issue is fixed. I wonder if there is some faster method to extract these data?
I think your script will work with a minor modification: load the data in chunks (you never need to load it all at once).