I'm interested in looking at the distribution of gravitational potentials for normal (baryonic) matter changing over time. That is, for each snapshot, computing the gravitational potential of each baryonic matter particle, plotting the distribution of (and calculating some basic statistics on) that, and examining how those change from snapshot to snapshot as particles cluster together and fall deeper into gravity wells.
If Illustris runs include an expansion factor due to "dark energy", is it possible to do this analysis without that?
Does this data already exist? If not, would it be reasonably easy to compute?
The Potential field is available for PartType0 (gas) for all baryonic runs. The units are described on the Data Specifications page. This is the same quantity computed by the code and used during the run to calculate gravitational accelerations - note that because we evolve the simulation in "comoving" coordinates (i.e. unchanging in the expanding universe), then I suspect this will already be what you want.
Thanks Dylan. Two questions:
(1) How do I get to the Potential field? When I look at the snapshot keys() in Illustris-3, I get [u'files', u'num_stars', u'filesize_snapshot', u'filesize_rockstar', u'redshift', u'url', u'num_gas', u'number', u'simulation', u'num_dm', u'num_groups_fof', u'num_groups_subfind', u'num_trmc', u'num_groups_rockstar', u'num_bhs', u'filesize_groupcat', u'subhalos']. Where under that is the potential?
(2) If I'm interested in luminous matter, shouldn't I be looking at PartType4 (stars and wind) rather than PartType0 (gas)?
I think you are listing here the fields from the API for a given snapshot, e.g. at
The Potential field is found only in the full snapshot files, e.g. download a full set of snapshot files like
and then (in each) at /PartType0/Potential/ for the potential at the positions of each gas cell, or /PartType4/Potential/ for the potential at the positions of each star/wind.
Ah, OK. So if I'm understanding you right, the Potential cannot be accessed through the web API, but only from downloaded files. I may need to buy a new hard disk, as even just Illustris-3 is 136 snapshots X 20 GB/snapshot = 2.7 TB, and I only have about 2 TB free at the moment.
This is right, Potential is "particle-level data" (just like e.g. the Velocity of individual star particles), which is only available in the actual snapshot files.
For the size, I just comment that the fact that 136 snapshots exist is rather arbitrary, you could simply download a subset (many of these are at z>8 anyways and not so useful).
I'm only interested in snapshots that are after the universe becomes transparent and that contain stars, so yes, I probably don't need the earliest ones. To start, I'm just going to download 135 and develop my scripts on that. But eventually, I would want to process all the relevant snapshots.