optim.IST

Contents

optim.IST#

class bartorch.optim.IST#

Bases:

Iterative soft thresholding (proximal gradient), looping ISTBlock.

Solves \(\min_x \tfrac12 \| A x - y \|^2 + g(x)\).

Parameters:

  • regularizers (Regularizer or ImplicitPrior, default=None) – The single term \(g\); its transform must be the identity.

  • maxiter (int, default=30) – Number of iterations.

  • step (float, default=0.95) – Step size \(\tau\) of the gradient step.

  • eigen (bool, default=False) – Divide the step by maxeigen() of the normal operator.

  • hogwild (bool, default=False) – Must be false: BART’s IST rejects it.

  • cclambda (float, default=0.0) – Weight of an identity added to the normal operator.

  • precond (LinearOperator, default=None) – Left preconditioner \(M\) from A.ishape to itself, positive definite (lsqr2_create’s precond_op); BART’s reconstructions use none.

Raises:

  • ValueError – Not exactly one term, or hogwild; at the call, a term whose transform is not the identity, such as TotalVariation.

  • TypeError – A term with auxiliary variables.

__call__()#

Solve for the image.

Parameters:

  • y (tensor) – Data of A.oshape, or (batch, *A.oshape); a batch is solved one item at a time, each as its own run.

  • A (LinearOperator) – Encoding operator. A BART-backed operator is applied inside the library; a Python-defined one is called back once per application.

  • x0 (tensor, default=None) – Starting image of A.ishape, or one per batch item; zero when omitted.

Returns:

Complex64 image of A.ishape, with y’s batch axis in front.

Return type:

torch.Tensor

Raises:

ValueError – y requires a gradient and a BART term has not been detached (detach()): its proximal operator has no backward pass.

Notes

The iterations are torch operations, so a solve is differentiated by unrolling. Step sizes and weights are fixed; bartorch.optim’s blocks and Unrolled expose them as parameters.

Contents