pallas::BoundedStepFunction Class Reference

A new candidate solution is generated between upper and lower bounds. Each degree of freedom in the candidate solution is at most step_size from the current solution. More...

#include <step_function.h>

Inheritance diagram for pallas::BoundedStepFunction:

Collaboration diagram for pallas::BoundedStepFunction:

Public Member Functions
	BoundedStepFunction (double step_size, const double *upper_bounds, const double *lower_bounds, unsigned int num_parameters)
	Constructor. More...
void	Step (double *x, unsigned int num_parameters)
	Modifies x in place to a new random position. More...
Public Member Functions inherited from pallas::StepFunction
virtual	~StepFunction ()
	Destructor.

Detailed Description

A new candidate solution is generated between upper and lower bounds. Each degree of freedom in the candidate solution is at most step_size from the current solution.

Parameters

step_size	double. Maximum magnitude a degree of freedom can be changed from its current position.
upper_bounds	double*. Pointer to the upper bounds on the candidate's variables.
lower_bounds	double*. Pointer to the lower bounds on the candidate's variables.
num_parameters	unsigned int. The number of parameters to modify in x.

Constructor & Destructor Documentation

pallas::BoundedStepFunction::BoundedStepFunction	(	double	step_size,
		const double *	upper_bounds,
		const double *	lower_bounds,
		unsigned int	num_parameters
	)

Constructor.

Parameters

step_size	double. Maximum magnitude a degree of freedom can be changed from its current position.
upper_bounds	double*. Pointer to the upper bounds on the candidate's variables.
lower_bounds	double*. Pointer to the lower bounds on the candidate's variables.
num_parameters	unsigned int. The number of parameters to modify in x. Should also be the number of constraints held within the upper and lower bounds.

Member Function Documentation

void pallas::BoundedStepFunction::Step ( double *  x, unsigned int  num_parameters  )

virtual

Modifies x in place to a new random position.

The maximum difference between the old and new position for each degree of freedom is determined by the step_size. If any degree of freedom exceeds the upper or lower bounds, it is adjusted to fall within the boundary.

Parameters

x	double*. The current state of the parameters. Modified in place to achieve new state.
num_parameters.	unsigned int. The number of parameters to modify in x.

Implements pallas::StepFunction.

---

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

• include/pallas/step_function.h
• src/pallas/step_function.cc