deepmr.optim.CGStep#

class deepmr.optim.CGStep(*args: Any, **kwargs: Any)[source]#

Conjugate Gradient method step.

This represents propagation through a single iteration of a CG algorithm; can be used to build unrolled architectures.

AHA#

Normal operator AHA = AH * A.

Type:

Callable | torch.Tensor

Ahy#

Adjoint AH of measurement operator A applied to the measured data y.

Type:

torch.Tensor

ndim#

Number of spatial dimensions of the problem. It is used to infer the batch axes. If AHA is a deepmr.linop.Linop operator, this is inferred from AHA.ndim and ndim is ignored.

Type:

int

tol#

Stopping condition. The default is None (run until niter).

Type:

float, optional

__init__(AHA, AHy, ndim=None, tol=None)[source]#

Methods

__init__(AHA, AHy[, ndim, tol])

check_convergence()

dot(s1, s2)

forward(input)