MultiAgentDecisionProcess
BayesianGameIdenticalPayoffSolver_T< JP > Class Template Reference

BayesianGameIdenticalPayoffSolver_T is an interface for solvers for Bayesian games with identical payoff. More...

#include <BayesianGameIdenticalPayoffSolver_T.h>

Inheritance diagram for BayesianGameIdenticalPayoffSolver_T< JP >:
[legend]

Public Member Functions

 BayesianGameIdenticalPayoffSolver_T (const boost::shared_ptr< const BayesianGameIdenticalPayoffInterface > &bg, size_t nrDesiredSolutions=1)
 (default) Constructor More...
 
virtual boost::shared_ptr
< JointPolicyDiscretePure
GetNewJpol () const
 this gives a implementation of GetNewJpol (specified in BayesianGameIdenticalPayoffSolver) More...
 
- Public Member Functions inherited from BayesianGameIdenticalPayoffSolver
void AddSolution (const JointPolicyPureVector &jp, double value)
 
void AddSolution (const JointPolicyPureVectorForClusteredBG &jp, double value)
 
void AddSolution (LIndex jpolIndex, double value)
 
 BayesianGameIdenticalPayoffSolver (const boost::shared_ptr< const BayesianGameIdenticalPayoffInterface > &bg, size_t nrDesiredSolutions=1)
 (default) Constructor More...
 
double Evaluate (const JointPolicyPureVector &jpolBG) const
 
double Evaluate (const JointPolicyPureVectorForClusteredBG &jpolBG) const
 
boost::shared_ptr< const
BayesianGameIdenticalPayoffInterface
GetBGIPI () const
 
double GetExpectedReward () const
 
const boost::shared_ptr
< JointPolicy
GetJointPolicy () const
 
const JointPolicyPureVectorGetJointPolicyPureVector () const
 
boost::shared_ptr< JPPVValuePairGetNextSolutionJPPV () const
 
boost::shared_ptr
< PartialJPDPValuePair
GetNextSolutionPJPDP () const
 
size_t GetNrDesiredSolutions () const
 
size_t GetNrFoundSolutions () const
 Gets the found number of solutions. More...
 
double GetPayoff () const
 
std::ofstream * GetResultsOFStream () const
 
std::ofstream * GetTimingsOFStream () const
 
bool GetWriteAnyTimeResults () const
 
bool IsEmptyJPPV () const
 
bool IsEmptyPJPDP () const
 
virtual bool IsExactSolver () const =0
 Methods should indicated whether they compute exact (optimal) solutions or not. More...
 
void PopNextSolutionJPPV ()
 
void PopNextSolutionPJPDP ()
 
void SaveSolution (const std::string &filename) const
 
void SetAnyTimeResults (bool turn_on, std::ofstream *results, std::ofstream *timings)
 Turns Anytime results on and of. More...
 
virtual void SetCBGlowerBound (double lowerbound)=0
 
virtual void SetCBGupperBound (double ub)=0
 
virtual void SetDeadline (double deadlineInSeconds)
 To limit the amount of time the solver uses. More...
 
void SetNrDesiredSolutions (size_t n)
 Gets the desired number of solutions to be returned. More...
 
std::string SoftPrintSolution () const
 
virtual double Solve ()=0
 The methods that performs the planning. More...
 
virtual ~BayesianGameIdenticalPayoffSolver ()
 Destructor. More...
 

Additional Inherited Members

- Protected Member Functions inherited from BayesianGameIdenticalPayoffSolver
virtual void CheckDeadline (const std::string &errorMessage) const
 Checks whether the deadline has expired. Throws EDeadline. More...
 
virtual void InitDeadline ()
 Should be called at the beginning of Solve(). More...
 

Detailed Description

template<class JP>
class BayesianGameIdenticalPayoffSolver_T< JP >

BayesianGameIdenticalPayoffSolver_T is an interface for solvers for Bayesian games with identical payoff.

The template argument JP represents the joint policy class the solver should return.

Constructor & Destructor Documentation

template<class JP>
BayesianGameIdenticalPayoffSolver_T< JP >::BayesianGameIdenticalPayoffSolver_T ( const boost::shared_ptr< const BayesianGameIdenticalPayoffInterface > &  bg,
size_t  nrDesiredSolutions = 1 
)
inline

(default) Constructor

takes a reference to the BG to be solved. nrDesiredSolutions is the number of solutions that the solver should return. I.e., if set higher than 1, it returns the k best found joint policies.

Member Function Documentation

template<class JP>
virtual boost::shared_ptr<JointPolicyDiscretePure> BayesianGameIdenticalPayoffSolver_T< JP >::GetNewJpol ( ) const
inlinevirtual

this gives a implementation of GetNewJpol (specified in BayesianGameIdenticalPayoffSolver)

Reimplemented from BayesianGameIdenticalPayoffSolver.

Referenced by BGIP_SolverMaxPlus< JP >::Solve(), and BGIP_SolverBFSNonIncremental< JP >::Solve().