---

GAManager Class Reference

#include <amygdala/gamanager.h>

List of all members.

Public Methods
	GAManager ()
virtual	~GAManager ()
void	Run (AmTimeInt maxSimTime, AmTimeInt maxRunTime)
virtual unsigned int	GetScore ()=0
virtual void	Setup (string uri)=0
int	Loop (string uri, AmTimeInt maxSimTime, AmTimeInt maxRunTime)
Network *	getNetwork ()
void	GenomeFormat (string format)
Protected Methods
void	finalize ()
Protected Attributes
Network *	NN
Genome *	genome
Static Protected Attributes
bool	run
bool	isManager
GAManager *	theGAManager

---

Detailed Description

Encapsulates the handling of new phenotypes

Author:

Rüdiger Koch

---

Member Function Documentation

void GAManager::finalize (    )  [protected]

return the results of the simulation

void GAManager::GenomeFormat (  string    format )

If we load from a file the genome doesn't get the format of the genome. Here we can supply it.

Network* GAManager::getNetwork (    )

return the network, simply to get hold of simTime

virtual unsigned int GAManager::GetScore (    )  [pure virtual]

A derived class must implement this function to return the score

int GAManager::Loop (  string    uri, AmTimeInt    maxSimTime, AmTimeInt    maxRunTime )

Fork a new child for each generation. We do this to avoid trouble with mem leaks, Amygdala initialization... Let the OS clean things up.

virtual void GAManager::Setup (  string    uri )  [pure virtual]

Set up the GA A derived class has to implement this class

---

Member Data Documentation

Network* GAManager::NN [protected]

the network

---

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

• gamanager.h

---