fdtdx.full_backward

Contents

fdtdx.full_backward#

fdtdx.full_backward(state, objects, config, key, record_detectors, reset_fields, start_time_step=0)[source]#

Perform full backward FDTD propagation from current state to start time.

Uses a while loop to repeatedly call backward() until reaching start_time_step. Leverages time-reversibility of Maxwell’s equations.

Parameters:

  • state (SimulationState) – Current simulation state tuple (time_step, arrays)

  • objects (ObjectContainer) – Container with simulation objects (sources, detectors, etc)

  • config (SimulationConfig) – Simulation configuration parameters

  • key (jax.Array) – JAX PRNG key for random operations

  • record_detectors (bool) – Whether to record detector states

  • reset_fields (bool) – Whether to reset fields after each step

  • start_time_step (int, optional) – Time step to propagate back to (default: 0)

Returns:

Final state after backward propagation

Return type:

SimulationState

Contents