pallas::SimulatedAnnealing::Options Struct Reference

#include <simulated_annealing.h>

Collaboration diagram for pallas::SimulatedAnnealing::Options:

Public Member Functions
void	set_step_function (scoped_ptr< StepFunction > &user_step_function)
	Convenience function for changing the default step function. More...

Public Attributes
CoolingSchedule::Options	cooling_schedule_options
GradientLocalMinimizer::Options	local_minimizer_options
scoped_ptr< StepFunction >	step_function
unsigned int	max_iterations
unsigned int	max_stagnant_iterations
unsigned int	dwell_iterations
double	minimum_cost
bool	polish_output
bool	is_silent
unsigned int	history_save_frequency

Detailed Description

Configurable options for modifying the default behaviour of the simulated annealing algorithm.

Member Function Documentation

void pallas::SimulatedAnnealing::Options::set_step_function ( scoped_ptr< StepFunction > &  user_step_function )

inline

Convenience function for changing the default step function.

This function simply swaps the scoped_ptr to the user defined StepFunction with the scoped_ptr to the StepFunction held within the pallas::SimulatedAnnealing::Options struct.

Parameters

user_step_function

pallas::scoped_ptr<pallas::StepFunction>. This function generates randomized candidate solutions based on the current position.

Member Data Documentation

CoolingSchedule::Options pallas::SimulatedAnnealing::Options::cooling_schedule_options

The annealing schedule to use. Must be one of ‘FAST’, ‘CAUCHY’ or ‘BOLTZMANN’. see pallas::CoolingSchedule.

unsigned int pallas::SimulatedAnnealing::Options::dwell_iterations

The number candidate solutions to generate before decreasing the temperature.

unsigned int pallas::SimulatedAnnealing::Options::history_save_frequency

Frequency to save the state of the system. Values will be appended to a HistorySeries contained in the optimization summary. Default is 0. If 0 then history is not saved. Otherwise, the state of the system will be appended to the series when i % history_save_frequency == 0. If there are a large number of iterations this can lead to a lot of data being stored in memory.

bool pallas::SimulatedAnnealing::Options::is_silent

Whether to log failure information relating the to global optimization algorithm using glog.

GradientLocalMinimizer::Options pallas::SimulatedAnnealing::Options::local_minimizer_options

Contains any changes to the default options for the local minimization algorithm. See the documentation for ceres::GradientProblemSolver::Options for relevant options

unsigned int pallas::SimulatedAnnealing::Options::max_iterations

Maximum number of simulated annealing iterations.

unsigned int pallas::SimulatedAnnealing::Options::max_stagnant_iterations

Maximum number sequential simulated annealing iterations allowed without finding a new global minimum.

double pallas::SimulatedAnnealing::Options::minimum_cost

User specified minimum cost. Minimization will halt if minimum_cost is reached.

bool pallas::SimulatedAnnealing::Options::polish_output

Whether the global minimum found through differential evolution should be subjected to a local minimization "polishing" step before returning the result.

scoped_ptr<StepFunction> pallas::SimulatedAnnealing::Options::step_function

Function that produces randomized candidate solutions.

---

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

• include/pallas/simulated_annealing.h