Caesar BH masses are incorrect

[DhruvZ]

I am on the most recent yt and caesar hashes and attempted to run caesar to generate a catalog. However the catalog's BH information is not accurate; the black hole masses do not make sense as they are all listed as less than 1 solar mass. Even if this is a conversion error between code mass and solar masses, it does not make sense because the masses would not match up with the black hole masses pulled from yt. Below is the script I used to demonstrate this.

import yt
import caesar
import numpy as np


ds = yt.load('/orange/narayanan/desika.narayanan/gizmo_runs/simba/m25n512/output/snapshot_305.hdf5')
proc = 16
# making caesar file
testObj = caesar.CAESAR(ds)
testObj.member_search(haloid = 'fof',blackholes=True,fof6d = True,nproc = proc)
testObj.save('test_caesar.hdf5')


# compare BH information with new caesar file
caes_test = caesar.load('test_caesar.hdf5')

yt_dat = ds.all_data()

# direct BH masses
bh_masses_yt = yt_dat["PartType5","Masses"]
bh_sorted_yt = np.sort(bh_masses_yt)[::-1] # large to small

# caesar stored BH masses
bh_masses_caes = [gal.masses['bh'].to('Msun') for gal in caes_test.galaxies]
bh_masses_caes = np.array(bh_masses_caes)

# print out comparable results - masses are not consistent with each other, even with potential factors of 10^10
print(bh_sorted_yt.to('Msun'))
print(bh_masses_caes)
# number of black holes also not close
print(len(bh_sorted_yt))
print(np.sum([gal.nbh for gal in caes_test.galaxies]))

[weiguangcui]

Hi @DhruvZ ,
Please update the pygadgetreader package (https://github.com/dnarayanan/pygadgetreader) and give it another try.
The issue with BH mass is not reading correctly by the previous pygadgetreader package. Please let me know if the problem persists.