guilhem_lavaux/JaxPM
0
0
Fork 1
mirror of https://github.com/DifferentiableUniverseInitiative/JaxPM.git synced 2025-04-07 20:30:54 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

ability to choose snapshots number with MultiHost script

Browse source
This commit is contained in:
Wassim KABALAN 2024-11-10 20:59:54 +01:00
parent ad4566620f
commit 12c74e2601
2 changed files with 318 additions and 292 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

581
notebooks/05-MultiHost_PM.ipynb
View file

File diff suppressed because one or more lines are too long

29
notebooks/05-MultiHost_PM.py
View file

@ -57,6 +57,12 @@ def parse_arguments():
help=
"Box size of the simulation as three values (e.g., 500.0 500.0 1000.0)."
)
parser.add_argument(
"-st",
"--snapshots",
type=int,
default=2,
help="Number of snapshots to save during the simulation.")
parser.add_argument("-H",
"--halo_size",
type=int,
@ -71,7 +77,7 @@ def parse_arguments():
return parser.parse_args()
def create_mesh_and_sharding(mesh_shape, pdims):
def create_mesh_and_sharding(pdims):
devices = create_device_mesh(pdims)
mesh = Mesh(devices, axis_names=('x', 'y'))
sharding = NamedSharding(mesh, P('x', 'y'))
@ -80,7 +86,7 @@ def create_mesh_and_sharding(mesh_shape, pdims):
@partial(jax.jit, static_argnums=(2, 3, 4, 5, 6))
def run_simulation(omega_c, sigma8, mesh_shape, box_size, halo_size,
solver_choice, sharding):
solver_choice,nb_snapshots, sharding):
k = jnp.logspace(-4, 1, 128)
pk = jc.power.linear_matter_power(
jc.Planck15(Omega_c=omega_c, sigma8=sigma8), k)
@ -115,7 +121,7 @@ def run_simulation(omega_c, sigma8, mesh_shape, box_size, halo_size,
dt0=0.01,
y0=jnp.stack([dx, p], axis=0),
args=cosmo,
saveat=SaveAt(ts=jnp.array([0.5, 1.0])),
saveat=SaveAt(ts=jnp.linspace(0.2, 1., nb_snapshots)),
stepsize_controller=stepsize_controller)
ode_fields = [
@ -132,18 +138,25 @@ def main():
box_size = args.box_size
halo_size = args.halo_size
solver_choice = args.solver
nb_snapshots = args.snapshots
mesh, sharding = create_mesh_and_sharding(mesh_shape, args.pdims)
sharding = create_mesh_and_sharding(args.pdims)
initial_conditions, lpt_displacements, ode_solutions, solver_stats = run_simulation(
0.25, 0.8, tuple(mesh_shape), tuple(box_size), halo_size,
solver_choice, sharding)
solver_choice, nb_snapshots, sharding)
if rank == 0:
os.makedirs("fields", exist_ok=True)
print(f"[{rank}] Simulation done")
print(f"Solver stats: {solver_stats}")
# Save initial conditions
initial_conditions_g = all_gather(initial_conditions)
if rank == 0:
print(f"[{rank}] Saving initial_conditions")
np.save("initial_conditions.npy", initial_conditions_g)
np.save("fields/initial_conditions.npy", initial_conditions_g)
print(f"[{rank}] initial_conditions saved")
del initial_conditions_g, initial_conditions
@ -151,7 +164,7 @@ def main():
lpt_displacements_g = all_gather(lpt_displacements)
if rank == 0:
print(f"[{rank}] Saving lpt_displacements")
np.save("lpt_displacements.npy", lpt_displacements_g)
np.save("fields/lpt_displacements.npy", lpt_displacements_g)
print(f"[{rank}] lpt_displacements saved")
del lpt_displacements_g, lpt_displacements
@ -160,7 +173,7 @@ def main():
sol_g = all_gather(sol)
if rank == 0:
print(f"[{rank}] Saving ode_solution_{i}")
np.save(f"ode_solution_{i}.npy", sol_g)
np.save(f"fields/ode_solution_{i}.npy", sol_g)
print(f"[{rank}] ode_solution_{i} saved")
del sol_g