grail::evolution::RouletteSampler< T > Class Template Reference

Provides a functionality of a fitness proportional roulette sampling of any objects that implement a GetFitness() method.
In particular it is used by an evolutionary algorithm (EA).
Because the implementation is optimized for usage in EA, the sampled elements are removed unless the method ends with 'WithReturn'. More...

#include <RouletteSampler.hh>

Public Member Functions
	RouletteSampler ()
	RouletteSampler - Constructor. Note that if you use a parameterless constructor, call Initialize() to provide elements for sampling. They must implement a public double GetFitness() method. More...
	RouletteSampler (const std::vector< std::unique_ptr< T >> &fitnessHavingElements)
	RouletteSampler - Constructor. More...
void	Initialize (const std::vector< std::unique_ptr< T >> &fitnessHavingElements)
	Sets the elements that can be sampled using this RouletteSampler. They must implement a public "double GetFitness()" method. More...
T *	Sample (std::mt19937_64 &randGen)
	Samples one element and removes it. More...
T *	SampleWithReturn (std::mt19937_64 &randGen)
	Samples one element without removing it. More...
std::vector< T * >	SampleMany (size_t count, std::mt19937_64 &randGen)
	Samples a given number of elements and removes them. More...
std::vector< T * >	SampleManyWithReturn (size_t count, std::mt19937_64 &randGen)
	Samples a given number of elements without removing them. More...
std::vector< size_t >	SampleManyIndices (size_t count, std::mt19937_64 &randGen)
	SampleManyIndices - samples a given number of indices of elements and removes them. More...

Detailed Description

template<class T>
class grail::evolution::RouletteSampler< T >

Provides a functionality of a fitness proportional roulette sampling of any objects that implement a GetFitness() method.
In particular it is used by an evolutionary algorithm (EA).
Because the implementation is optimized for usage in EA, the sampled elements are removed unless the method ends with 'WithReturn'.

Parameters

T	- the type T must have a public method ef "double GetFitness()"

Constructor & Destructor Documentation

RouletteSampler() [1/2]

template<class T >

grail::evolution::RouletteSampler< T >::RouletteSampler

inline

RouletteSampler - Constructor. Note that if you use a parameterless constructor, call Initialize() to provide elements for sampling.
They must implement a public double GetFitness() method.

RouletteSampler() [2/2]

template<class T >

grail::evolution::RouletteSampler< T >::RouletteSampler ( const std::vector< std::unique_ptr< T >> &  fitnessHavingElements )

inline

RouletteSampler - Constructor.

Parameters

fitnessHavingElements

- a vector of elements which will be sampled.

Member Function Documentation

Initialize()

template<class T >

void grail::evolution::RouletteSampler< T >::Initialize ( const std::vector< std::unique_ptr< T >> &  fitnessHavingElements )

inline

Sets the elements that can be sampled using this RouletteSampler. They must implement a public "double GetFitness()" method.

Parameters

fitnessHavingElements

- a vector of elements which will be sampled.

Sample()

template<class T >

T * grail::evolution::RouletteSampler< T >::Sample ( std::mt19937_64 &  randGen )

inline

Samples one element and removes it.

Parameters

randGen

- random numbers generator to use.

Returns

the sampled element.

SampleMany()

template<class T >

std::vector< T * > grail::evolution::RouletteSampler< T >::SampleMany ( size_t  count, std::mt19937_64 &  randGen  )

inline

Samples a given number of elements and removes them.

Parameters

count	- the number of elements to sample.
randGen	- random numbers generator to use.

Returns

the sampled elements.

SampleManyIndices()

template<class T >

std::vector< size_t > grail::evolution::RouletteSampler< T >::SampleManyIndices ( size_t  count, std::mt19937_64 &  randGen  )

inline

SampleManyIndices - samples a given number of indices of elements and removes them.

Parameters

count	- the number of indices to sample.
randGen	- random numbers generator to use.

Returns

the sampled indices of elements.

SampleManyWithReturn()

template<class T >

std::vector< T * > grail::evolution::RouletteSampler< T >::SampleManyWithReturn ( size_t  count, std::mt19937_64 &  randGen  )

inline

Samples a given number of elements without removing them.

Parameters

count	- the number of elements to sample.
randGen	- random numbers generator to use.

Returns

the sampled elements.

SampleWithReturn()

template<class T >

T * grail::evolution::RouletteSampler< T >::SampleWithReturn ( std::mt19937_64 &  randGen )

inline

Samples one element without removing it.

Parameters

randGen

- random numbers generator to use.

Returns

the sampled element.

---

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

• GrailEvolution/RouletteSampler.hh