prepare

experimental.network_dynamics.prepare(
    dynamics,
    solver,
    t0=0.0,
    t1=1.0,
    dt=0.1,
    n_nodes=1,
    noise=None,
    externals=None,
)

Prepare a bare dynamics model for simulation using Diffrax.

Lightweight alternative to the Network-based prepare() for uncoupled simulations with adaptive time stepping. Coupling inputs are set to zero.

Parameters

Name	Type	Description	Default
dynamics	AbstractDynamics	Dynamics model instance (e.g. JansenRit, ReducedWongWang)	required
solver	DiffraxSolver	Diffrax integration method (Tsit5, Dopri5, etc.)	required
t0	float	Start time, by default 0.0	0.0
t1	float	End time, by default 1.0	1.0
dt	float	Initial step size (dt0) for adaptive controller, by default 0.1	0.1
n_nodes	int	Number of uncoupled nodes, by default 1	1
noise	AbstractNoise	Noise process (e.g. AdditiveNoise, MultiplicativeNoise)	None
externals	dict	External inputs as {name: AbstractExternalInput}	None

Returns

Name	Type	Description
solve_function	Callable	Pure JAX function: solve_function(config) -> results
config	Bunch	Configuration PyTree with dynamics params and initial state