SequentialReconstructor#

class snake.toolkit.reconstructors.pysap.SequentialReconstructor(*, nufft_kwargs=<factory>, max_iter_per_frame=15, optimizer='pogm', wavelet='db4', threshold='sure', nufft_backend='gpunufft', density_compensation='pipe', restart_strategy=RestartStrategy.WARM, compute_backend='cupy')[source]#

Bases: BaseReconstructor

Use a sequential Reconstruction.

Parameters:
  • max_iter_frame – Number of iteration to allow per frame.

  • optimizer (str) – Optimizer name, available are pogm and fista.

  • threshold (float | str) – Threshold value for the wavelet regularisation.

  • nufft_kwargs (dict[str, Any])

  • max_iter_per_frame (int)

  • wavelet (str)

  • nufft_backend (str)

  • density_compensation (str | bool)

  • restart_strategy (RestartStrategy)

  • compute_backend (str)

Methods

__init__

reconstruct

Reconstruct with Sequential.

setup

Set up the reconstructor.

Attributes

compute_backend

density_compensation

log

Get a logger.

max_iter_per_frame

nufft_backend

optimizer

restart_strategy

threshold

wavelet

nufft_kwargs

setup(sim_conf)[source]#

Set up the reconstructor.

Parameters:

sim_conf (SimConfig)

Return type:

None

reconstruct(data_loader, sim_conf)[source]#

Reconstruct with Sequential.

Parameters:
Return type:

ndarray

property log: Logger#

Get a logger.