Creating a catalog of galaxy SFHs

Bhavin Joshi
  • 7 Jun '23


I am attempting to create a catalog of SFHs of Illustris-TNG galaxies. I'm using the answer provided here ( and other scripts elsewhere on the TNG website.
I started with the lower resolution TNG 300-3 simulation to develop my code but this sim appears to have few valid subhalos. For the z=0 snapshot, while the total subhalo count is almost 4e5 it seems like most of these don't have any information in their corresponding dictionaries.

Here's a MWE:

import illustris_python as il

datadir = '/Volumes/Joshi_external_HDD/IllustrisTNG/TNG300-3/'
basePath = datadir + 'output'
snapnum = 99

gcount = il.groupcat.loadSubhalos(basePath, snapnum)
total_subhalos = gcount['count']
print('Total subhalos:', total_subhalos)

empty_subhalo_count = 0
for i in range(total_subhalos):
    subhalo_id = i
    subhalo_stars = il.snapshot.loadSubhalo(basePath, snapnum, subhalo_id, 'stars')
        initial_stellar_masses = subhalo_stars['GFM_InitialMass']
    except KeyError:
        print('Empty subhalo...')
        empty_subhalo_count += 1

print('Total empty subhalos:', empty_subhalo_count)

This returns an "empty" subhalo count of about 3.3e5 (so about 6e4 valid subhalos). After a couple other selection cuts prior to extracting SFHs, my final sample count is about 1e4.
Am I interpreting this correctly? If so, are there more subhalos in the full resolution run of TNG300 so that I can build a larger sample of galaxy SFHs?

Thanks very much,
Bhavin Joshi

Dylan Nelson
  • 7 Jun '23

TNG300-3 is extremely low resolution, it is possible that most subhalos do not form stars. So, it may not be the best test case.

I would suggest instead to test with TNG50-4, which is even smaller (in particle count), but has better (i.e. more realistic) resolution.

Bhavin Joshi
  • 7 Jun '23

Thanks Dylan. I'll try TNG50-4

  • Page 1 of 1