import mosaic
from .pipeline import Pipeline
__all__ = ['ProcessWavelets', 'ProcessTraces',
'ProcessGlobalGradient', 'ProcessModelIteration']
# TODO Default configuration of pipelines should be better defined
[docs]@mosaic.tessera
class ProcessWavelets(Pipeline):
"""
Default pipeline to process wavelets before running the forward problem.
**Default steps:**
- ``filter_wavelets``
"""
def __init__(self, steps=None, no_grad=False, **kwargs):
steps = steps or []
super().__init__(steps, no_grad=no_grad, **kwargs)
[docs]@mosaic.tessera
class ProcessTraces(Pipeline):
"""
Default pipeline to process modelled and observed before running the functional.
**Default steps:**
- ``mute_traces``
- ``filter_traces``
- ``norm_per_shot``
"""
def __init__(self, steps=None, no_grad=False, **kwargs):
steps = steps or []
if kwargs.pop('mute_traces', True):
steps.append('mute_traces')
steps.append('filter_traces')
norm_per_shot = kwargs.pop('norm_per_shot', True)
norm_per_trace = kwargs.pop('norm_per_trace', False)
if norm_per_trace:
steps.append('norm_per_trace')
elif norm_per_shot:
steps.append('norm_per_shot')
super().__init__(steps, no_grad=no_grad, **kwargs)
[docs]@mosaic.tessera
class ProcessGlobalGradient(Pipeline):
"""
Default pipeline to process the global gradient before updating the variable.
**Default steps:**
- ``mask``
- ``smooth_field``
- ``norm_field``
"""
def __init__(self, steps=None, no_grad=True, **kwargs):
steps = steps or []
mask = kwargs.pop('mask', True)
if mask:
steps.append('mask')
smooth = kwargs.pop('smooth', True)
if smooth:
steps.append('smooth_field')
norm = kwargs.pop('norm', True)
if norm:
steps.append('norm_field')
super().__init__(steps, no_grad=no_grad, **kwargs)
[docs]@mosaic.tessera
class ProcessModelIteration(Pipeline):
"""
Default pipeline to process the model after each iteration.
**Default steps:**
- ``clip``
"""
def __init__(self, steps=None, no_grad=True, **kwargs):
steps = steps or []
steps.append('clip')
super().__init__(steps, no_grad=no_grad, **kwargs)