idealized Arepo simulation support

[dnelson86]

We have so far focused mostly on cosmological simulations.

Let us also check minimal support for idealized simulations (e.g. much different units, different field conventions possible, etc).

e.g. isolated galaxy disk in arepo/examples/ (from public version of the code).

[arkordt]

I would like to support this issue and offer support for extending and generalizing scida in that regard. Here is a related issue that I encountered:

summary of the arepo setup

• cosmological simulation of a 1 (Mpc/h)^3 box
• these arepo modules produce output data
  • TRACER_MC as particle type 2
  • SGChem with radiative and chemical properties saved in particle type 0
  • supernova remnants maintained in particle type 3
  • SPRAI radiative transfer with sink particles as particle type 5
• basically, I am using an identical simulation setup as in Magg et al. 2022

After defining some additional units for quantities, opening the snapshot works fine. While following the scida documentation regarding visualization with holoviews and datashader, generating the dataframe including all fields does not work. Excerpt from the error report

...

File ~/miniforge3/envs/astro/lib/python3.10/site-packages/scida/fields.py:504, in FieldContainer.get_dataframe(self, fields)
    502     idim = int(k[i:])
    503     k = k.split(k[i:])[0]
--> 504 v = self[k]
    505 assert v.ndim <= 2  # cannot support more than 2 here...
    506 if idim is not None:

...

File ~/miniforge3/envs/astro/lib/python3.10/site-packages/scida/customs/arepo/dataset.py:1658, in Temperature(arrs, ureg, **kwargs)
   1655     m_p = m_p * ureg.g
   1656     k_B = k_B * ureg.erg / ureg.K
-> 1658 xe = arrs[\"ElectronAbundance\"]
   1659 u_internal = arrs[\"InternalEnergy\"]
   1661 mu = 4 / (1 + 3 * xh + 4 * xh * xe) * m_p

...

File ~/miniforge3/envs/astro/lib/python3.10/site-packages/scida/fields.py:661, in FieldContainer._getitem(self, key, force_derived, update_dict, evaluate_recipe)
    659     return field
    660 else:
--> 661     raise KeyError(\"Unknown field '%s'\" % key)

KeyError: \"Unknown field 'ElectronAbundance'\""

Reading through src/scida/customs/arepo/dataset.py, I think that the ArepoSnapshot class should be much more general, e.g.

• lines 90-97: other than PartType0 and PartType1, there should not be any other default alias
• lines 1640-1664: PartType0->ElectronAbundance does not exist but could be inferred from ChemicalAbundances array