First Steps Walkthrough

First, start a command-line ipython session, or a Jupyter notebook. Then import the library

import temet

Simulation Selection

Most analysis is based around a “simulation parameters” object (typically called sim below), which specifies the simulation and snapshot of interest, among other details. You can then select a simulation and snapshot from a known list, e.g. the TNG100-2 simulation of IllustrisTNG at redshift two:

sim = temet.sim(res=910, run='tng', redshift=2.0)

Or the TNG50-1 simulation at redshift zero, by its short-hand name

sim = temet.sim(run='tng50-1', redshift=0.0)

Note that if you would like to select a simulation which is not in the pre-defined list of known simulations, i.e. a simulation you have run yourself, then you can specify it simply by path

sim = temet.sim('/home/user/sims/sim_run/', redshift=0.0)

In all cases, the redshift or snapshot number is optionally used to pre-select the particular snapshot of interest. You can also specify the overall simulation without this, for example

sim = temet.sim('tng300-1')

Loading Data

Once a simulation and snapshot is selected you can load the corresponding data. For example, to load one or more particular fields from the group catalogs

subs = sim.groupCat(fieldsSubhalos=['SubhaloMass','SubhaloPos'])
sub_masses_logmsun = sim.units.codeMassToLogMsun( subs['SubhaloMass'] )

To load particle-level data from the snapshot itself

gas_pos = sim.snapshotSubset('gas', 'pos')
star_masses = sim.snapshotSubsetP('stars', 'mass')
dm_vel_sub10 = sim.snapshotSubset('dm', 'vel', subhaloID=10)

In addition to shorthand names for fields such as “pos” (mapping to “Coordinates”), many custom fields at both the particle and group catalog level are defined. Note that snapshotSubsetP() is the parallel (multi-threaded) version, and will be significantly faster. Loading data can also be done with shorthands, for example

subs = sim.subhalos('mstar_30pkpc')
x = sim.gas('cellsize_kpc')

fof10 = sim.halo(10) # all fields
sub_sat1 = sim.subhalo( fof10['GroupFirstSub']+1 ) # all fields

Exploratory Plots for Galaxies

The various plotting functions in plot.cosmoGeneral are designed to be as general and automatic as possible. They are idea for a quick look or for exploring trends in the objects of the group catalogs, i.e. galaxies (subhalos).

Let’s examine a classic observed galaxy scaling relation: the correlation between gas-phase metallicity, and stellar mass, the “mass-metallicity relation” (MZR).

sim = temet.sim(run='tng100-1', redshift=0.0)

temet.plot.cosmoGeneral.quantMedianVsSecondQuant(sim, 'Z_gas', 'mstar_30pkpc')

This produces a PDF figure named medianQuants_TNG100-1_Z_gas_mstar_30pkpc_cen.pdf in the current working directory. It shows the mass-metallicity relation of TNG100 galaxies at \(z=0\), and looks like this:

_images/first_steps_medianQuants_1.png

We can enrich the plot in a number of ways, both by tweaking minor aesthetic options, and by including additional information from the simulation. For example, we will shift the x-axis bounds, and also include individual subhalos as colored points, coloring based on gas fraction:

sim = temet.sim(run='tng100-1', redshift=0.0)

temet.plot.cosmoGeneral.quantMedianVsSecondQuant(sim, 'Z_gas', 'mstar_30pkpc',
  xlim=[8.0, 11.5], scatterColor='fgas2')

This produces the following figure, which highlights how lower mass galaxies have high gas fractions of nearly unity, i.e. \(M_{\rm gas} \gg M_\star\), and that gas fraction slowly decreasing with stellar mass until \(M_\star \sim 10^{10.5} M_\odot\). At this point, the overall gas metallicity turns over and starts to decrease, as indicated by the black median line. Gas fractions also drop rapidly, reaching \(f_{\rm gas} \sim 10^{-4}\) before starting to slowly rise again. This feature marks the onset of galaxy quenching due to supermassive black hole feedback.

_images/first_steps_medianQuants_2.png

Once you add a custom calculation for a new property of subhalos, i.e. compute a value which isn’t available by default, you can use the same plotting routines to understand how it varies across the galaxy population, and correlates with other galaxy properties.

Exploratory Plots for Snapshots

Similarly, plot.general provides general plotting routines focused on snapshots, i.e. particle-level data. These are also then suitable for non-cosmological simulations.

Functionality includes 1D histograms, 2D distributions, median relations, and radial profiles.

For example, we could plot the traditional 2D “phase diagram” of density versus temperature. However, we can also use any (known) quantity on either axis. Furthermore, while color can represent the distribution of mass, it can also be used to show the value of a third particle/cell property, in each pixel. Let’s look at the relationship between gas pressure and magnetic field strength at \(z=0\):

sim = temet.sim(run='tng100-1', redshift=0.0)

temet.plot.general.plotPhaseSpace2D(sim, 'gas', xQuant='pres', yQuant='bmag')
_images/first_steps_phase2D_1.png

For cosmological simulations, we can also look at particle/cell properties for one or more (stacked) halos. For example, the relationship between (halocentric) radial velocity and (halocentric) distance, for all dark matter particles within the tenth most massive halo of TNG50-1 at \(z=2\):

sim = temet.sim(run='tng50-1', redshift=2.0)
haloIDs = [9]

opts = {'xlim':[-0.6,0.3], 'ylim':[-800,600], 'clim':[-4.7,-2.3], 'ctName':'inferno'}
temet.plot.general.plotPhaseSpace2D(sim, 'dm', xQuant='rad_rvir', yQuant='vrad', haloIDs=haloIDs, **opts)
_images/first_steps_phase2D_2.png

Here we see individual gravitationally bound substructures (subhalos) within the halo as bright vertical features.

Visualizing a Halo and its Galaxy

TODO.

Computing a Custom Post-processing Catalog

So far we have been exclusively exploring and visualizing existing data – either properties which are directly available in the catalogs or snapshots (e.g. galaxy stellar mass, gas cell magnetic field strength), or which can be easily derived from them (e.g. dark matter particle radial velocity with respect to its parent halo).

Instead, we may be interested to compute a new physical quantity of interest for each object in the catalog.

We typically refer to such results as “post-processing catalogs”, “supplementary catalogs”, or “auxiliary catalogs”.

TODO.