Add support for tracer particles

[pgrete]

Alright, I'm done with the cleanup,i.e., I

• separated the general from the problem specific pieces
• added appropriate callbacks
• added a test
• added doc

It'd be great if you, @BenWibking and @mfournier01 could review from the "all the basic bases are covered" point of view.
There's some stuff left I/we may want to address (in the near future) but I think it's cleaner to address those in separate smaller PRs.
I'm particularly interested in @mfournier01 feedback from the "documentation" and "is it clear what's going on" point of view as you're a new developer that will have to modify the code very soon.
So I'd be happy to provide more info in the code/docs already in this PR to make things clearer where required.

Future cleanup items:

• Allow for 2D sims
• Add additional/other integrators/advection methods
• Find origin of tiny drift
• Migrate to SwarmPacks
• Allow more flexiblity wrt tracked fields (e.g., not make/calc prims by default)

Python analysis/plotting example:

import matplotlib.pyplot as plt
import numpy as np
import sys
sys.path.insert(
    1,
    "/home/pgrete/src/athenapk/external/parthenon"
    + "/scripts/python/packages/parthenon_tools/parthenon_tools",
)

try:
    import phdf
except ModuleNotFoundError:
    print("Couldn't find module to read Parthenon hdf5 files.")

data = phdf.phdf("../run-turbulence/parthenon.prim.00010.phdf")
tracers = data.GetSwarm("tracers")
xs = tracers.x
ys = tracers.y
zs = tracers.z
ids = tracers.Get("id")

fig = plt.figure()
ax = fig.add_subplot(projection='3d')

ax.scatter(xs, ys, zs,s=1,c=ids)#,cmap="plasma")

results in

with lookback times available in

data.Params["tracers/t_lookback"]

from parameters in an output block like

<parthenon/output2>
file_type  = hdf5       # Binary data dump
variables   = prim   # variables to be output
dt         = 0.1        # time increment between outputs
id         = prim
single_precision_output = true

swarms = tracers
tracers_variables = id, x, y, z, rho, s, sdot

and tracers enabled by an additional block as

<tracers>
enabled = true
num_tracers_per_cell = 0.125

[mfournier01]

Left a few comments for clarifications of what the code is doing but I haven't found any typo. The documentation is pretty clear.

[mfournier01]

Not yet clear to me: why do we need only one tacklist per region in that case?

[pgrete]

I'm actually not sure if this is 100% required, but as long as we're running with a single host thread this should not make any difference.

[BenWibking]

Minor comments inline.

Also, I added a plotting script in scripts/ based on the example you provided in the docs. I also uploaded an input file that included a <tracers> block, since it looked like none of the existing problems had one.

[BenWibking]

Are particles included in the XDMF output? If not, I think it would be helpful to be explicit that this is the case.

[pgrete]

They're currently not. I added the info.

[pgrete]

Actually, I take that back. They can be written to xdmf but it has to be manually activated.
I now extended the example input file and doc.

[pgrete]

[BenWibking]

Looks nice!

tracers.mp4

[BenWibking]

LGTM