slope::SortedL1Norm Class Reference

Class representing the Sorted L1 Norm. More...

#include <sorted_l1_norm.h>

Public Member Functions
double	eval (const Eigen::VectorXd &beta, const Eigen::ArrayXd &lambda) const
	Evaluates the Sorted L1 Norm.
Eigen::MatrixXd	prox (const Eigen::VectorXd &beta, const Eigen::ArrayXd &lambda) const
	Computes the proximal operator of the Sorted L1 Norm.
double	dualNorm (const Eigen::VectorXd &a, const Eigen::ArrayXd &lambda) const
	Computes the dual norm of a vector.
double	dualNorm (const Eigen::VectorXd &a, const Eigen::ArrayXd &lambda, double denominator_floor) const
	Computes the dual norm with a numerical denominator floor.

Detailed Description

Class representing the Sorted L1 Norm.

Definition at line 15 of file sorted_l1_norm.h.

Member Function Documentation

dualNorm() [1/2]

double slope::SortedL1Norm::dualNorm	(	const Eigen::VectorXd &	a,
		const Eigen::ArrayXd &	lambda
	)		const

Computes the dual norm of a vector.

Parameters

a	The vector.
lambda	The regularization weights.

Returns

The dual norm.

dualNorm() [2/2]

double slope::SortedL1Norm::dualNorm	(	const Eigen::VectorXd &	a,
		const Eigen::ArrayXd &	lambda,
		double	denominator_floor
	)		const

Computes the dual norm with a numerical denominator floor.

This overload is used for restricted working-set convergence checks, where treating a tiny positive penalty as exactly zero can prevent the working set from expanding. Full dual values use the unfloored overload.

Parameters

a	The vector whose dual norm is computed.
lambda	The regularization weights.
denominator_floor	Minimum cumulative penalty denominator.

Returns

The numerically regularized dual norm.

eval()

double slope::SortedL1Norm::eval	(	const Eigen::VectorXd &	beta,
		const Eigen::ArrayXd &	lambda
	)		const

Evaluates the Sorted L1 Norm.

Parameters

beta	The beta parameter.
lambda	The regularization weights.

Returns

The evaluation result.

prox()

Eigen::MatrixXd slope::SortedL1Norm::prox	(	const Eigen::VectorXd &	beta,
		const Eigen::ArrayXd &	lambda
	)		const

Computes the proximal operator of the Sorted L1 Norm.

Parameters

beta	The beta parameter.
lambda	The regulariation weights.

Returns

The proximal operator result.

---

The documentation for this class was generated from the following file:

• include/slope/sorted_l1_norm.h