proxtoolbox.proxoperators package¶

This package contains the ProxOperator abstract class and various concrete classes.

Submodules¶

proxtoolbox.proxoperators.ADMM_prox module¶

class proxtoolbox.proxoperators.ADMM_prox.ADMM_Context(experiment)[source]¶

Bases: object

Class containing data and helper functions for ADMM prox operators and ADMM related classes

class proxtoolbox.proxoperators.ADMM_prox.Prox_primal_ADMM_indicator(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator, proxtoolbox.proxoperators.ADMM_prox.ADMM_Context

Projecting onto physical (object) domain constraints, adapted for the ADMM algorithm This is an approximate projector onto a ball around the data determined by the Kullback-Leibler divergence, as appropriate for Poisson noise. The analysis of such approximate projections was studied in D.R.Luke, “Local Linear Convergence of Approximate Projections onto Regularized Sets”, Nonlinear Analysis, 75(2012):1531–1546.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on Oct 4, 2017.

eval(u, prox_index=None)[source]¶

Projection subroutine for projecting onto physical (object) domain constraints, adapted for the ADMM algorithm.

Parameters:	u : cell Input data to be projected prox_idx : int, optional Index of this prox operator
Returns:	xprime : cell The projection

class proxtoolbox.proxoperators.ADMM_prox.Approx_Pphase_FreFra_ADMM_Poisson(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator, proxtoolbox.proxoperators.ADMM_prox.ADMM_Context

Prox operator for computing the update to the auxilliary variables in the ADMM scheme subject to constraints in the image domain (the intensity measurements in the Fresnel/Fraunhofer domain.

The projection onto the data constraints is an approximate projector onto a ball around the data determined by the Kullback-Leibler divergence, as appropriate for Poisson noise. The analysis of such approximate projections was studied in D.R.Luke, “Local Linear Convergence of Approximate Projections onto Regularized Sets”, Nonlinear Analysis, 75(2012):1531–1546.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on Oct 4, 2017.

eval_helper(u, j, proj_scale=1.0, proj_const=0.0)[source]¶

eval(u, prox_idx=None)[source]¶

Projection onto the data constraints

Parameters:	u : cell Image in the physical domain to be projected. assumed here to be a cell, the first element of which project onto the amplitude constraints prox_idx : int, optional Index of this prox operator
Returns:	u_image : cell The projection in the same format as u.

proxtoolbox.proxoperators.Approx_Pphase_FreFra_Poisson module¶

class proxtoolbox.proxoperators.Approx_Pphase_FreFra_Poisson.Approx_Pphase_FreFra_Poisson(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Prox operator implementing projection onto Fourier magnitude constraints in the near field or far field with Poisson noise.

This is an approximate projector onto a ball around the data determined by the Kullback-Leibler divergence, as appropriate for Poisson noise. The analysis of such approximate projections was studied in D.R.Luke, “Local Linear Convergence of Approximate Projections onto Regularized Sets”, Nonlinear Analysis, 75(2012):1531–1546.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) around Oct 3, 2017.

eval(u, prox_idx=None)[source]¶

Projection onto Fourier magnitude constraints in the near field or far field with Poisson noise.

Parameters:	u : array-like Function in the physical domain to be projected prox_idx : int, optional Index of this prox operator
Returns:	u_new : array_like The projection in the physical (time) domain in the same format as u.

proxtoolbox.proxoperators.CT_prox module¶

class proxtoolbox.proxoperators.CT_prox.CT_ProxOperator(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Base class for CT prox operators

class proxtoolbox.proxoperators.CT_prox.P_hyperplane(experiment)[source]¶

Bases: proxtoolbox.proxoperators.CT_prox.CT_ProxOperator

Projection onto hyperplane

eval(u, prox_idx=None)[source]¶

Projects a point u onto a hyperplane given by the prox_idx-th row of matrix A and b[prox_idx]

Parameters:	u : array-like point to be projected (vector) prox_idx : int Index of this prox operator
Returns:	y : array-like The projection (vector).

class proxtoolbox.proxoperators.CT_prox.P_sequential_hyperplane_odd(experiment)[source]¶

Bases: proxtoolbox.proxoperators.CT_prox.CT_ProxOperator

Hyperplane projection subroutine of the ART method for tomographic reconstruction of a density profile.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on May 29, 2012.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : array-like point to be projected (vector) prox_idx : int, optional Index of this prox operator
Returns:	v : array-like The projection (vector).

class proxtoolbox.proxoperators.CT_prox.P_sequential_hyperplane_even(experiment)[source]¶

Bases: proxtoolbox.proxoperators.CT_prox.CT_ProxOperator

Hyperplane projection subroutine of the ART method for tomographic reconstruction of a density profile.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on May 29, 2012.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : array-like point to be projected (vector) prox_idx : int, optional Index of this prox operator
Returns:	v : array-like The projection (vector).

proxtoolbox.proxoperators.P_Amod module¶

class proxtoolbox.proxoperators.P_Amod.P_Amod(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection onto support constraints

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on June 1, 2017.

eval(u, prox_idx=None)[source]¶

Projects the input data onto support constraints.

Parameters:	u : ndarray Function in the physical domain to be projected prox_idx : int, optional Index of this prox operator
Returns:	p_A : ndarray The projection in the physical (time) domain. Leaves the phase alone on the support, resets amplitude to 1 and sets amplitude to 1 with 0 phase outside support

proxtoolbox.proxoperators.P_CDP module¶

class proxtoolbox.proxoperators.P_CDP.P_CDP(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Prox operator that projects onto Fourier magnitude constraints with nonunitary, but invertible masks.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on Sept. 27, 2016.

eval(u, prox_index=None)[source]¶

Projection onto Fourier magnitude constraints with nonunitary, but invertible masks

Parameters:	u : ndarray Input data to be projected prox_idx : int, optional Index of this prox operator
Returns:	u_new : ndarray The projection

proxtoolbox.proxoperators.P_M module¶

class proxtoolbox.proxoperators.P_M.P_M(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection onto Fourier magnitude constraints

Initialization

Parameters:	config : dict - Dictionary containing the problem configuration. It must contain the following mapping: ‘M’ : array_like - non-negative real FOURIER DOMAIN CONSTRAINT

eval(u, prox_idx=None)[source]¶

Applies the proxoperator P_M

Parameters:	u : array_like - function in the physical domain to be projected
Returns:	array_like - [p_M,phat_M] ,where p_M = the projection IN THE PHYSICAL (time) DOMAIN and phat_M = projection IN THE FOURIER DOMAIN

class proxtoolbox.proxoperators.P_M.P_M_masked(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_M.P_M

Projection onto Fourier magnitude constraints, leaving alone points masked by ‘fmask’

Initialization

Parameters:	experiment: experiment class

eval(u, prox_idx=None)[source]¶

Applies the proxoperator P_M_masked

Parameters:	u : array_like - function in the physical domain to be projected
Returns:	array_like - p_M: the projection IN THE PHYSICAL (time) DOMAIN

class proxtoolbox.proxoperators.P_M.Approx_P_M(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_M.P_M

eval(u, prox_idx=None)[source]¶

Applies the proxoperator P_M

Parameters:	u : array_like - function in the physical domain to be projected
Returns:	array_like - [p_M,phat_M] ,where p_M = the projection IN THE PHYSICAL (time) DOMAIN and phat_M = projection IN THE FOURIER DOMAIN

class proxtoolbox.proxoperators.P_M.Approx_P_M_masked(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_M.P_M_masked

eval(u, prox_idx=None)[source]¶

Applies the proxoperator P_M_masked

Parameters:	u : array_like - function in the physical domain to be projected
Returns:	array_like - p_M: the projection IN THE PHYSICAL (time) DOMAIN

proxtoolbox.proxoperators.P_SP module¶

class proxtoolbox.proxoperators.P_SP.P_SP(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection onto nonnegativity and support constraints

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on May 23, 2002.

eval(u, prox_idx=None)[source]¶

Projects the input data onto nonnegativity and support constraints.

Parameters:	u : ndarray Function in the physical domain to be projected prox_idx : int, optional Index of this prox operator
Returns:	p_SP : ndarray The projection in the physical (time) domain

proxtoolbox.proxoperators.P_Sparsity module¶

class proxtoolbox.proxoperators.P_Sparsity.P_Sparsity(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection subroutine for projecting onto a sparsity constraint

Initialization

Parameters:	config : dict - Dictionary containing the problem configuration. It must contain the following mappings: ‘sparsity_parameter’ : int

eval(u, prox_idx=None)[source]¶

Parameters:	u : array_like - Input, array to be projected
Returns:	p_Sparsity : array_like, the projection

class proxtoolbox.proxoperators.P_Sparsity.P_Sparsity_real(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_Sparsity.P_Sparsity

Projection subroutine for projecting onto a combined sparsity and realness constraint

order: realness, sparsity

Initialization

Parameters:	config : dict - Dictionary containing the problem configuration. It must contain the following mappings: ‘sparsity_parameter’ : int

eval(u, prox_idx=None)[source]¶

Parameters:	u : array_like - Input, array to be projected
Returns:	p_Sparsity : array_like, the projection

class proxtoolbox.proxoperators.P_Sparsity.P_Sparsity_Symmetric(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_Sparsity.P_Sparsity

Projection subroutine for projecting onto a combined sparsity and symmetry constraint

order: symmetry, sparsity

eval(u, prox_idx=None)[source]¶

Parameters:	u : array_like - Input, array to be projected
Returns:	p_Sparsity : array_like, the projection

class proxtoolbox.proxoperators.P_Sparsity.P_Sparsity_Symmetric_real(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_Sparsity.P_Sparsity_Symmetric

Projection subroutine for projecting onto a combined symmetry, sparsity and realness constraint

order: realness, symmetry, sparsity

eval(u, prox_idx=None)[source]¶

Parameters:	u : array_like - Input, array to be projected
Returns:	p_Sparsity : array_like, the projection

class proxtoolbox.proxoperators.P_Sparsity.P_Sparsity_Superpixel(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_Sparsity.P_Sparsity

Apply sparsity on superpixels, i.e. on the binned array

eval(u, prox_idx=None)[source]¶

Parameters:	u : array_like - Input, array to be projected
Returns:	p_Sparsity : array_like, the projection

class proxtoolbox.proxoperators.P_Sparsity.P_Sparsity_Superpixel_real(experiment)[source]¶

Bases: proxtoolbox.proxoperators.P_Sparsity.P_Sparsity_Superpixel

Apply real-valued sparsity on superpixels, i.e. on the binned array

eval(u, prox_idx=None)[source]¶

Parameters:	u : array_like - Input, array to be projected
Returns:	p_Sparsity : array_like, the projection

proxtoolbox.proxoperators.P_amp_support module¶

class proxtoolbox.proxoperators.P_amp_support.P_amp_support(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection onto support constraints

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on July 26, 2011.

eval(u, prox_idx=None)[source]¶

Projects the input data onto support constraints. ‘abs_illumination’ is the object domain constraint (0-1 indicator function)

Parameters:	u : ndarray or a list of ndarray objects Input data to be projected prox_idx : int, optional Index of this prox operator
Returns:	p_amp : ndarray or a list of ndarray objects The projection

proxtoolbox.proxoperators.P_avg module¶

class proxtoolbox.proxoperators.P_avg.P_avg(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection onto the diagonal, averaging the input

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on Feb. 19, 2018.

eval(u, prox_index=None)[source]¶

Averages the input u. This is a projection onto onto the diagonal of a product space, but this routine dow not use that structure. The only difference with P_Diag is that the output is not copied onto the diagonal of the implicit product spcace.

Parameters:	u : ndarray a list of ndarray objects Input data to be projected prox_idx : int, optional Index of this prox operator
Returns:	u_avg : ndarray or a list of ndarray objects The projection, the average.

proxtoolbox.proxoperators.P_diag module¶

class proxtoolbox.proxoperators.P_diag.P_diag(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection onto the diagonal of a product space

eval(u, prox_idx=None)[source]¶

Projects the input data onto the diagonal of the product space given by its dimensions.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on July 26, 2011.

Parameters:	u : ndarray or a list of ndarray objects Input data to be projected prox_idx : int, optional Index of this prox operator
Returns:	u_diag : ndarray or a list of ndarray objects The projection

proxtoolbox.proxoperators.Prox_product_space module¶

class proxtoolbox.proxoperators.Prox_product_space.Prox_product_space(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Prox operator that projects onto constraints arranged in a product space.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on Oct 26, 2017.

eval(u, prox_index=None)[source]¶

Projection subroutine onto constraints arranged in a product space.

Parameters:	u : ndarray or a list of ndarray objects Input data to be projected prox_idx : int, optional Index of this prox operator
Returns:	u_new : ndarray or a list of ndarray objects The projection, in the product space.

proxtoolbox.proxoperators.propagators module¶

class proxtoolbox.proxoperators.propagators.PropagatorFFTn(experiment)[source]¶

Bases: object

Basic propagator which uses the n-dimensional fourier transform, transforming over all axes by default

Intialize propagator

Parameters:	experiment – experiment class, can be used to pass options as attributes of the class

eval(u, **kwargs)[source]¶

class proxtoolbox.proxoperators.propagators.InvPropagatorFFTn(experiment)[source]¶

Bases: proxtoolbox.proxoperators.propagators.PropagatorFFTn

Basic inverse propagator which uses the inverse n-dimensional fourier transform, transforming over all axes by default

Intialize propagator

Parameters:	experiment – experiment class, can be used to pass options as attributes of the class

eval(u, **kwargs)[source]¶

class proxtoolbox.proxoperators.propagators.PropagatorFFT2(experiment)[source]¶

Bases: proxtoolbox.proxoperators.propagators.PropagatorFFTn

Basic propagator which uses the two-dimensional fourier transform

Intialize propagator

Parameters:	experiment – experiment class, can be used to pass options as attributes of the class

class proxtoolbox.proxoperators.propagators.InvPropagatorFFT2(experiment)[source]¶

Bases: proxtoolbox.proxoperators.propagators.PropagatorFFT2

Basic inverse propagator which uses the inverse two-dimensional fourier transform

Intialize propagator

Parameters:	experiment – experiment class, can be used to pass options as attributes of the class

eval(u, **kwargs)[source]¶

class proxtoolbox.proxoperators.propagators.Propagator_FreFra(experiment)[source]¶

Bases: proxtoolbox.proxoperators.propagators.Propagator_FreFra_Base

Propagator for near field or far field Fourier measurements.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) around Jan 23, 2019.

eval(u, prox_idx=None)[source]¶

Propagation function for near field or far field Fourier measurements.

Parameters:	u : array_like Function in the physical domain to be projected prox_idx : int, optional Index of the prox operator calling this method
Returns:	u_hat : array_like the propagated field at the measurement plane

class proxtoolbox.proxoperators.propagators.InvPropagator_FreFra(experiment)[source]¶

Bases: proxtoolbox.proxoperators.propagators.Propagator_FreFra_Base

Inverse propagator for near field or far field Fourier measurements

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) around Jan 23, 2019.

eval(p_Mhat, prox_idx=None)[source]¶

Inverse propagation function for near field or far field Fourier measurements.

Parameters:	p_Mhat : array_like Function in the measurement plane to be propagated back to the physical plane prox_idx : int, optional Index of the prox operator calling this method
Returns:	u_new : array_like Propagated field at the object plane

proxtoolbox.proxoperators.proxoperator module¶

class proxtoolbox.proxoperators.proxoperator.ProxOperator(experiment)[source]¶

Bases: object

Generic interface for prox operators

Initialization method for a concrete instance

Parameters:	experiment : instance of Experiment class Experiment object that will use this prox operaror

eval(u, prox_idx=None)[source]¶

Applies a prox operator to some input data

Parameters:	u : ndarray or a list of ndarray objects Input data to be projected prox_idx : int, optional Index of this prox operator
Returns:	ndarray or a list of ndarray objects Result of the application of the prox operator onto the input data

proxtoolbox.proxoperators.proxoperator.magproj(constr, u)[source]¶

Projection operator onto a magnitude constraint.

Inexact, but stable implementation of magnitude projection. See LukeBurkeLyon, SIREV 2002.

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on July 24, 2001.

Parameters:	constr : array_like A nonnegative array that is the magnitude constraint. u : array_like The function to be projected onto constr (can be complex).
Returns:	array_like The projection.

proxtoolbox.proxoperators.ptychographyProx module¶

class proxtoolbox.proxoperators.ptychographyProx.ExplicitPtychographyProxOperator(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Base class for Ptychography explicit prox opertaors

class proxtoolbox.proxoperators.ptychographyProx.Explicit_ptychography_scan_farfield_probe_ptwise(experiment)[source]¶

Bases: proxtoolbox.proxoperators.ptychographyProx.ExplicitPtychographyProxOperator

The probe update for the PHeBIE algorithm

Ref.: “Proximal Heterogeneous Block Implicit-Explicit Method and Application to Blind Ptychographic Diffraction Imaging”, R. Hesse, D. R. Luke, S. Sabach and M. K. Tam, SIAM J. on Imaging Sciences, 8(1):426–457 (2015).

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on 16th February 2019.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : Cell Iterate. Assumes u[0] is the probe, u[1] is the object, u[2] is the block of variables (product space) satisfying the intensity measurements prox_idx : int, optional Index of this prox operator
Returns:	u_new : Cell The projection.

class proxtoolbox.proxoperators.ptychographyProx.Explicit_ptychography_scan_farfield_object_ptwise(experiment)[source]¶

Bases: proxtoolbox.proxoperators.ptychographyProx.ExplicitPtychographyProxOperator

The object update for the PHeBIE algorithm

Ref.: “Proximal Heterogeneous Block Implicit-Explicit Method and Application to Blind Ptychographic Diffraction Imaging”, R. Hesse, D. R. Luke, S. Sabach and M. K. Tam, SIAM J. on Imaging Sciences, 8(1):426–457 (2015).

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on 16th February 2019.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : Cell Iterate. Assumes u[0] is the probe, u[1] is the object, u[2] is the block of variables (product space) satisfying the intensity measurements prox_idx : int, optional Index of this prox operator
Returns:	u_new : Cell The projection.

class proxtoolbox.proxoperators.ptychographyProx.Explicit_ptychography_image(experiment)[source]¶

Bases: proxtoolbox.proxoperators.ptychographyProx.ExplicitPtychographyProxOperator

This is the argument of the projection in Step 3 of Algorithm 4.1 of “Proximal Heterogeneous Block Implicit-Explicit Method and Application to Blind Ptychographic Diffraction Imaging”, R. Hesse, D. R. Luke, S. Sabach and M. K. Tam, SIAM J. on Imaging Sciences, 8(1):426–457 (2015).

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on 16th February 2019.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : Cell Iterate. Assumes u[0] is the probe, u[1] is the object, u[2] is the block of variables (product space) satisfying the intensity measurements prox_idx : int, optional Index of this prox operator
Returns:	u_new : Cell The projection.

class proxtoolbox.proxoperators.ptychographyProx.P_ptychography_scan_farfield_probe(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

The probe update for the PHeBIE algorithm Ref.: “Proximal Heterogeneous Block Implicit-Explicit Method and Application to Blind Ptychographic Diffraction Imaging”, R. Hesse, D. R. Luke, S. Sabach and M. K. Tam, SIAM J. on Imaging Sciences, 8(1):426–457 (2015).

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on 16th February 2019.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : Cell Iterate. assumes u is the probe (i.e., only part u[0] of the whole collection of PHeBIE variable blocks) prox_idx : int, optional Index of this prox operator
Returns:	u_new : Cell The projection.

class proxtoolbox.proxoperators.ptychographyProx.P_supp_amp_band(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

The object update for the PHeBIE algorithm Ref.: “Proximal Heterogeneous Block Implicit-Explicit Method and Application to Blind Ptychographic Diffraction Imaging”, R. Hesse, D. R. Luke, S. Sabach and M. K. Tam, SIAM J. on Imaging Sciences, 8(1):426–457 (2015).

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on 16th February 2019.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : Cell Iterate. assumes u is the object (i.e., only part u[1] of the whole collection of PHeBIE variable blocks) prox_idx : int, optional Index of this prox operator
Returns:	u_new : Cell The projection.

class proxtoolbox.proxoperators.ptychographyProx.Prox_ptychography_ptwise_diag(experiment)[source]¶

Bases: proxtoolbox.proxoperators.ptychographyProx.ExplicitPtychographyProxOperator

The object update for the PHeBIE algorithm

Ref.: “Proximal Heterogeneous Block Implicit-Explicit Method and Application to Blind Ptychographic Diffraction Imaging”, R. Hesse, D. R. Luke, S. Sabach and M. K. Tam, SIAM J. on Imaging Sciences, 8(1):426–457 (2015).

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on 16th February 2019.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : Cell Cell of arrays in the domain to be projected. Assumes u[1] is the object, u[2] is the block of variables (product space) satisfying the intensity measurements prox_idx : int, optional Index of this prox operator
Returns:	u_new : Cell The projection.

class proxtoolbox.proxoperators.ptychographyProx.Prox_ptychography_product_space(experiment)[source]¶

Bases: proxtoolbox.proxoperators.ptychographyProx.ExplicitPtychographyProxOperator

Projection subroutine onto constraints arranged in a product space

Based on Matlab code written by Russell Luke (Inst. Fuer Numerische und Angewandte Mathematik, Universitaet Gottingen) on Oct 26, 2017.

eval(u, prox_idx=None)[source]¶

Evaluate this prox operator

Parameters:	u : Cell Cell of arrays in the domain to be projected. Assumes u[1] is the object, u[2] is the block of variables (product space) satisfying the intensity measurements. prox_idx : int, optional Index of this prox operator
Returns:	u_new : Cell The projection, a cell of arrays in the product space

proxtoolbox.proxoperators.sudokuProx module¶

class proxtoolbox.proxoperators.sudokuProx.SudokuProx(experiment)[source]¶

Bases: proxtoolbox.proxoperators.proxoperator.ProxOperator

Projection onto the given entries in a Sudoku problem

class proxtoolbox.proxoperators.sudokuProx.ProjGiven(experiment)[source]¶

Bases: proxtoolbox.proxoperators.sudokuProx.SudokuProx

Projection onto the given entries in a Sudoku problem

eval(u, prox_idx=None)[source]¶

Applies the prox operator to the input data

Parameters:	u : array-like Iterate. prox_idx : int, optional Index of this prox operator
Returns:	A : array-like The projection.

class proxtoolbox.proxoperators.sudokuProx.ProjSquare(experiment)[source]¶

Bases: proxtoolbox.proxoperators.sudokuProx.SudokuProx

Projection onto the given entries in a Sudoku problem

eval(u, prox_idx=None)[source]¶

Applies the prox operator to the input data

Parameters:	u : array-like Iterate. prox_idx : int, optional Index of this prox operator
Returns:	Q : array-like The projection.

class proxtoolbox.proxoperators.sudokuProx.ProjColumn(experiment)[source]¶

Bases: proxtoolbox.proxoperators.sudokuProx.SudokuProx

Projection onto the given entries in a Sudoku problem

eval(u, prox_idx=None)[source]¶

Applies the prox operator to the input data

Parameters:	u : array-like Iterate. prox_idx : int, optional Index of this prox operator
Returns:	C : array-like The projection.

class proxtoolbox.proxoperators.sudokuProx.ProjRow(experiment)[source]¶

Bases: proxtoolbox.proxoperators.sudokuProx.SudokuProx

Projection onto the given entries in a Sudoku problem

eval(u, prox_idx=None)[source]¶

Applies the prox operator to the input data

Parameters:	u : array-like Iterate. prox_idx : int, optional Index of this prox operator
Returns:	C : array-like The projection.