pagmo::algorithm::pade Class Reference

Parallel Decomposition (PaDe) More...

#include <pade.h>

Inheritance diagram for pagmo::algorithm::pade:

Public Types
enum	weight_generation_type { RANDOM =0, GRID =1, LOW_DISCREPANCY =2 }
	Mechanism used to generate the weight vectors. More...

Public Member Functions
	pade (int gen=10, unsigned int max_parallelism=1, pagmo::problem::decompose::method_type=pagmo::problem::decompose::BI, const pagmo::algorithm::base &=pagmo::algorithm::jde(100), population::size_type=8, weight_generation_type=LOW_DISCREPANCY, const fitness_vector &=std::vector< double >())
	Constructor. More...
	pade (const pade &)
	Copy constructor. Performs a deep copy. Necessary as a pointer to a base algorithm is here contained.
base_ptr	clone () const
	Clone method.
void	evolve (population &) const
	Evolve implementation. More...
std::string	get_name () const
	Algorithm name.
std::vector< fitness_vector >	generate_weights (const unsigned int, const unsigned int) const
	Generates the weights used in the problem decomposition. More...
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

Parallel Decomposition (PaDe)

This class implement a multi-objective optimization algorithm based on parallel decomposition. For each element of the population a different single objective problem is generated using a decomposition method. Those single-objective problems are thus solved in parallel. At the end of the evolution the population is set as the best individual for each single-objective problem.

PaDe assumes all the objectives need to be minimized.

Author

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

Dario Izzo (dario.nosp@m..izz.nosp@m.o@gma.nosp@m.il.c.nosp@m.om)

Definition at line 52 of file pade.h.

Member Enumeration Documentation

enum pagmo::algorithm::pade::weight_generation_type

Mechanism used to generate the weight vectors.

Enumerator
RANDOM	Weights are generated uniformly at random on the simplex.
GRID	Weights are generated on a uniform grid layed down on the simplex.
LOW_DISCREPANCY	Weights are generated on the simplex with low-discrepancy.

Definition at line 56 of file pade.h.

Constructor & Destructor Documentation

pagmo::algorithm::pade::pade	(	int	gen = 10,
		unsigned int	threads = 1,
		pagmo::problem::decompose::method_type	method = pagmo::problem::decompose::BI,
		const pagmo::algorithm::base &	solver = pagmo::algorithm::jde(100),
		population::size_type	T = 8,
		weight_generation_type	weight_generation = LOW_DISCREPANCY,
		const fitness_vector &	z = std::vector<double>()
	)

Constructor.

Constructs a PaDe algorithm

Parameters

[in]	gen	Number of generations to evolve.
[in]	threads	the amounts of threads that will be used
[in]	method	the decomposition method to use (Weighted, Tchebycheff or BI)
[in]	solver	the algorithm to solve the single objective problems.
[in]	T	the size of the population on each subproblem (must be an even number)
[in]	weight_generation	the method to generate the weight vectors (RANDOM, GRID or LOW-DISCREPANCY)
[in]	z	the reference point used for decomposition (with Tchebycheff and BI)

Exceptions

value_error

if gen is negative, weight_generation is not sane

See also

pagmo::problem::decompose::method_type

Definition at line 69 of file pade.cpp.

Member Function Documentation

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

virtual

Evolve implementation.

Run the PaDe 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 217 of file pade.cpp.

std::vector< fitness_vector > pagmo::algorithm::pade::generate_weights	(	const unsigned int	n_f,
		const unsigned int	n_w
	)		const

Generates the weights used in the problem decomposition.

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

Parameters

[in]	n_f	diemension of the fitness space
[in]	n_w	number of weights to be produced

Definition at line 140 of file pade.cpp.

std::string pagmo::algorithm::pade::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 379 of file pade.cpp.

---

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

• pade.h
• pade.cpp