CompositeSignalModel#
- class CompositeSignalModel(*args)[source]#
Class that sequentially calls the signal-process building blocks. This offers a general way to stack and exchange arbitrary building blocks without changing the simulation calling signature
- Parameters:
args – All signal modules that need to be concatenated.
Methods:
__call__(signal_tensor[, segment_index])Consecutively calls submodules which the signal change to the passed in tensor.
unstack_repetitions(simulation_result)Uses the dimension expansion information from the sub modules to unstack the simulated tensor of k-space, samples or images.
update()Calls update function of all sub-modules and records the overall expansion factor
Attributes:
Returns the total expansion factor (factor by which the #repetitions axis grows)
returns the name of modules with expansion factors > 1 in the order that is used in unstack_repetitions:
- __call__(signal_tensor, segment_index=0, **kwargs)[source]#
Consecutively calls submodules which the signal change to the passed in tensor.
- Parameters:
kwargs – dictionary of tensors containing all required quantities. Is forwarded to BaseSignalModule.
signal_tensor (Tensor) –
segment_index (int | Tensor | None) –
- unstack_repetitions(simulation_result)[source]#
Uses the dimension expansion information from the sub modules to unstack the simulated tensor of k-space, samples or images. In both cases the second axis (index=1) is assumed to represent the stacked repetitions. :type simulation_result:
Union[Tensor,ndarray] :param simulation_result: (-1, [noise], samples) :rtype:Tensor:return: k-space (…, [noise], samples)- Parameters:
simulation_result (Tensor | ndarray) –
- Return type:
Tensor
- property expected_number_of_repetitions: Tensor#
Returns the total expansion factor (factor by which the #repetitions axis grows)