differential_evolution.h File Reference

#include <cfloat>
#include "pallas/history_concept.h"
#include "pallas/types.h"
#include "pallas/internal/crossover_strategy.h"
#include "pallas/internal/mutation_strategy.h"
#include "pallas/internal/shuffler.h"
#include "pallas/internal/state.h"

Include dependency graph for differential_evolution.h:

Go to the source code of this file.

Classes
class	pallas::DifferentialEvolution
	Minimizes an objective function by continuously evolving a population of candidate solutions. More...
struct	pallas::DifferentialEvolution::Options
struct	pallas::DifferentialEvolution::Summary
	Contains a summary of the optimization. More...
struct	pallas::DifferentialEvolution::HistoryOutput
	Stores information about the state of the system for at a given iteration number. More...

Functions
void	pallas::Solve (const DifferentialEvolution::Options &options, const GradientProblem &problem, double *parameters, DifferentialEvolution::Summary *summary)
	Helper function which avoids going through the interface of the pallas::DifferentialEvolution class. More...
void	pallas::dump (const DifferentialEvolution::HistoryOutput &h, HistoryWriter &writer)
	Dumps the system state contained in the history output into the stream contained by the writer. More...

Detailed Description

Author

Ryan Latture

Date

5-18-15

This file contains a C++ implementation of the differential algorithm due to Storn and Price: Differential Evolution - a Simple and Efficient Heuristic for Global Optimization over Continuous Spaces.

This code is based on the SciPy implementation of differential evolution found in scipy.optimize and relies on the local optimization algorithms in the Google Ceres project.

Function Documentation

void pallas::dump	(	const DifferentialEvolution::HistoryOutput &	h,
		HistoryWriter &	writer
	)

Dumps the system state contained in the history output into the stream contained by the writer.

Parameters

h	DifferentialEvolution::HistoryOutput. State of the system for a specific iteration.
writer	HistoryWriter. Object responsible for writing the history output to a stream.

Here is the call graph for this function:

void pallas::Solve	(	const DifferentialEvolution::Options &	options,
		const GradientProblem &	problem,
		double *	parameters,
		DifferentialEvolution::Summary *	summary
	)

Helper function which avoids going through the interface of the pallas::DifferentialEvolution class.

The specified options are used to setup a differential evolution instance which is then used to minimize the GradientProblem. The optimal solution is stored in parameters and a summary of the global optimization can be found in summary.

Parameters

options	pallas::DifferentialEvolution::Options. Options used to configure the optimization.
problem	pallas::GradientProblem. The problem to optimize.
parameters	double*. The starting point for further optimization.
summary	DifferentialEvolution::Summary*. Summary instance to store the optimization details.

Here is the call graph for this function: