Skip to main content

iteration_terminators

This module contains functions that determine when to stop the iteration process. The functions are used as arguments in the reconstruct function in the algorithm.py module. The functions are simple and can be easily replaced with custom functions.

All Iteration Terminator functions can only accept one positional argument of type: TerminatorInputs. This structure is created and passed in by the reconstruction algorithm.

Class: TerminatorInputs​

fpm_py.iteration_terminators.TerminatorInputs:
    ...

Dataclass for the inputs to the iteration terminator functions. All iteration terminator functions must accept this dataclass as the only positional argument.

Attributes: object (torch.Tensor): The object array. Not used in this simple terminator. wave_fourier (torch.Tensor): The current Fourier domain of the wave. None on the first iteration. wave_fourier_new (torch.Tensor): The new Fourier domain of the wave. None on the first iteration. i (int): The current iteration number.

Function: iter_ceil​

fpm_py.iteration_terminators.iter_ceil(
    inputs: TerminatorInputs,
    max_iters: int
):
    ...

Simple iteration terminator that stops after a given number of iterations.

Args:​

  • inputs (TerminatorInputs): The inputs to the terminator function.
  • max_iters (int): The maximum number of iterations to run.

Returns:​

  • bool: True if the iteration number is greater than or equal to the maximum number of iterations