astronomix.option_classes.simulation_params module

Runtime simulation parameters.

Unlike the simulation configuration, these parameters can be changed without triggering a recompilation, and the simulation can be differentiated with respect to them (CFL number, gas constants, viscosity, end time, …). The per-module parameter containers are bundled in here too.

class astronomix.option_classes.simulation_params.FixedBoundaryState(x: FixedBoundaryState1D = (Array([], shape=(0,), dtype=float32), Array([], shape=(0,), dtype=float32)), y: FixedBoundaryState1D = (Array([], shape=(0,), dtype=float32), Array([], shape=(0,), dtype=float32)), z: FixedBoundaryState1D = (Array([], shape=(0,), dtype=float32), Array([], shape=(0,), dtype=float32)))

Bases: NamedTuple

Per-axis fixed boundary states used when a boundary is FIXED_BOUNDARY.

x: FixedBoundaryState1D

Alias for field number 0

y: FixedBoundaryState1D

Alias for field number 1

z: FixedBoundaryState1D

Alias for field number 2

class astronomix.option_classes.simulation_params.FixedBoundaryState1D(left_state: Array = Array([], shape=(0,), dtype=float32), right_state: Array = Array([], shape=(0,), dtype=float32))

Bases: NamedTuple

The prescribed left/right states for a single axis with FIXED_BOUNDARY.

left_state: Array

Left state of shape (num_variables,).

right_state: Array

Right state of shape (num_variables,).

class astronomix.option_classes.simulation_params.SimulationParams(C_cfl: float = 0.4, gravitational_constant: float = 1.0, gravitational_potential: array = Array([], shape=(0,), dtype=float32), viscosity: float = 0.0, thermal_conductivity: float = 0.0, isothermal_sound_speed: float = 1.0, gamma: float = 1.6666666666666667, minimum_density: float = 1e-14, minimum_pressure: float = 1e-14, positivity_max_velocity: float = 50.0, dt_max: float = inf, t_start: float = 0.0, t_end: float = 0.2, snapshot_timepoints: array = Array([0.], dtype=float32), fixed_boundary_state: FixedBoundaryState = ((Array([], shape=(0,), dtype=float32), Array([], shape=(0,), dtype=float32)), (Array([], shape=(0,), dtype=float32), Array([], shape=(0,), dtype=float32)), (Array([], shape=(0,), dtype=float32), Array([], shape=(0,), dtype=float32))), turbulent_forcing_params: TurbulentForcingParams = (0.02, 50.0, 2.0, 1.0, 4.0, 1.0), wind_params: WindParams = (0.0, 0.0, 100000.0), cosmic_ray_params: CosmicRayParams = (0.0, 0.1), cooling_params: CoolingParams = (0.76, 0.02, 10000.0, (1.0, 1.0, 100000000.0)), neural_net_force_params: NeuralNetForceParams = (None,), cnn_mhd_corrector_params: CNNMHDconfig = (False, None))

Bases: NamedTuple

Different from the simulation configuration, the simulation parameters do not require recompilation when changed. The simulation can be differentiated with respect to them.

C_cfl: float

The Courant-Friedrichs-Lewy number, a factor in the time step calculation.

cnn_mhd_corrector_params: CNNMHDconfig

The parameters of the CNN MHD corrector module.

cooling_params: CoolingParams

The parameters of the cooling module.

cosmic_ray_params: CosmicRayParams

Cosmic ray parameters

dt_max: float

The maximum time step.

fixed_boundary_state: FixedBoundaryState

The fixed boundary state if the boundary type of the specific boundary is set to FIXED_BOUNDARY.

gamma: float

The adiabatic index of the gas.

gravitational_constant: float

Gravitational constant.

gravitational_potential: array

External, static gravitational potential evaluated at the cell centers of the grid (without ghost cells). Only used when config.gravity_config.external_potential is True; it is added on top of the self-gravity potential in _compute_total_potential and padded to the ghost-cell-extended shape of a state field internally.

isothermal_sound_speed: float

The isothermal sound speed used when config.equation_of_state is ISOTHERMAL. NOTE: CURRENTLY ONLY IMPLEMENTED FOR FINITE DIFFERENCE MODE.

minimum_density: float

Minimum allowed density. NOTE: CURRENTLY ONLY USED IN FINITE DIFFERENCE MODE IF positivity protection is active.

minimum_pressure: float

Minimum allowed pressure. NOTE: CURRENTLY ONLY USED IN FINITE DIFFERENCE MODE IF positivity protection is active.

neural_net_force_params: NeuralNetForceParams

The parameters of the neural network force module.

positivity_max_velocity: float

Velocity ceiling applied to cells fixed by the REDISTRIBUTE positivity mode (mirrors HOW-MHD velpmx1). Only used when a positivity mode is POSITIVITY_REDISTRIBUTE.

snapshot_timepoints: array

Snapshot timepoints

t_end: float

The final time of the simulation.

t_start: float

The initial (clock) time of the simulation. The time loop starts integrating from here and the snapshot grid spans [t_start, t_end]. Defaults to 0.0; set to a checkpoint’s time to resume a run (see astronomix.setup_helpers.restart_from_latest_checkpoint).

thermal_conductivity: float

Constant thermal conductivity kappa in the conductive energy source div(kappa grad T) (config.thermal_conduction). T is taken from the ideal-gas relation T = p / rho (code units, R = 1). NOTE: CURRENTLY ONLY IMPLEMENTED FOR FINITE DIFFERENCE MODE.

turbulent_forcing_params: TurbulentForcingParams

The parameters of the turbulent forcing module.

viscosity: float

Dynamic or kinematic viscosity depending on the viscosity_type in SimulationConfig.

wind_params: WindParams

The parameters of the stellar wind module.