cg#

mrinufft.extras.cg(operator: FourierOperatorBase, kspace_data: NDArray, damp: float = 0.0, x0: NDArray | None = None, x_init: NDArray | None = None, max_iter: int = 10, tol: float = 0.0001, progressbar: bool | tqdm = True, callback: Callable | None = None)[source]#

Perform conjugate gradient (CG) optimization for image reconstruction.

The image is updated using the gradient of a data consistency term, and a velocity vector is used to accelerate convergence.

Parameters:

  • $base_params

  • tol (float) – Tolerance for converge check.

  • $returns

References

https://en.m.wikipedia.org/wiki/Nonlinear_conjugate_gradient_method

Note

This function uses numpy for all CPU arrays, and cupy for all on-gpu array. It will convert all its array argument to the respective array library. The outputs will be converted back to the original array module and device.

See also

  • This function is part of the optimizer registry. You can find other registered functions in this registry below:

Key

Function

"lsqr"

lsqr(operator, kspace_data, damp=0.0, atol=1e-06, btol=1e-06, conlim=100000000.0, max_iter=100, x0=None, x_init=None, callback=None, progressbar=True)

"lsmr"

lsmr(operator, kspace_data, damp=0.0, atol=1e-06, btol=1e-06, conlim=100000000.0, max_iter=100, x0=None, x_init=None, callback=None, progressbar=True)

"cg"

cg(operator, kspace_data, damp=0.0, x0=None, x_init=None, max_iter=10, tol=0.0001, progressbar=True, callback=None)