pagmo::algorithm::firefly Class Reference

The Firefly algorithm. More...

#include <firefly.h>

Inheritance diagram for pagmo::algorithm::firefly:

Public Member Functions
	firefly (int gen=1, double alpha=0.01, double beta=1.0, double gamma=0.8)
	Constructor. More...
base_ptr	clone () const
	Clone method.
void	evolve (population &) const
	Evolve implementation. More...
std::string	get_name () const
	Algorithm name.
Public Member Functions inherited from pagmo::algorithm::base
	base ()
	Default constructor. More...
virtual	~base ()
	Trivial destructor. More...
std::string	human_readable () const
	Return human readable representation of the algorithm. More...
void	set_screen_output (const bool p)
	Setter-Getter for protected m_screen_output data. More...
bool	get_screen_output () const
	Gets screen output. More...
void	reset_rngs (const unsigned int) const
	Resets the seed of the internal rngs using a user-provided seed. More...

Protected Member Functions
std::string	human_readable_extra () const
	Extra human readable algorithm info. More...

Friends
class	boost::serialization::access

Additional Inherited Members
Protected Attributes inherited from pagmo::algorithm::base
bool	m_screen_output
	Indicates to the derived class whether to print stuff on screen.
rng_double	m_drng
	Random number generator for double-precision floating point values.
rng_uint32	m_urng
	Random number generator for unsigned integer values.
unsigned int	m_fevals
	A counter for the number of function evaluations.

Detailed Description

The Firefly algorithm.

The firefly algorithm (FA) is a metaheuristic algorithm, inspired by the flashing behaviour of fireflies.

At each call of the evolve method a number of function evaluations equal to gen * pop.size() * pop.size() is performed.

NOTE: when called on mixed-integer problems Firefly treats the integer part as fixed and optimizes the continuous part.

The algorithm used is the one provided in the first paper attached with 2 differences:

is calculated with the adaptive method explained in the second paper attached (Equation 3) The attactiveness is calculated as

These modifications make the algorithm perform better on all the test problems we experimented.

See also

http://arxiv.org/abs/1003.1466

http://www.springerlink.com/content/au3w21311g465007/

Author

Andrea Mambrini (andre.nosp@m.a.ma.nosp@m.mbrin.nosp@m.i@gm.nosp@m.ail.c.nosp@m.om)

Definition at line 60 of file firefly.h.

Constructor & Destructor Documentation

pagmo::algorithm::firefly::firefly	(	int	gen = 1,
		double	alpha = 0.01,
		double	beta = 1.0,
		double	gamma = 0.8
	)

Constructor.

Allows to specify in detail all the parameters of the algorithm.

Parameters

[in]	gen	number of iterations.
[in]	alpha	define the width of the random vector
[in]	beta	define the maximum attractiveness
[in]	gamma	define the absorption coefficent

Exceptions

value_error

if number of iterations is negative or alpha, beta and gamma are not in [0,1]

Definition at line 51 of file firefly.cpp.

Member Function Documentation

void pagmo::algorithm::firefly::evolve ( population &  pop ) const

virtual

Evolve implementation.

Run the Firefly algorithm for the number of generations specified in the constructors.

Parameters

[in,out]

pop

input/output pagmo::population to be evolved.

Implements pagmo::algorithm::base.

Definition at line 78 of file firefly.cpp.

std::string pagmo::algorithm::firefly::human_readable_extra ( ) const

protectedvirtual

Extra human readable algorithm info.

Will return a formatted string displaying the parameters of the algorithm.

Reimplemented from pagmo::algorithm::base.

Definition at line 212 of file firefly.cpp.

---

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

• firefly.h
• firefly.cpp