FactoredDecPOMDPDiscreteInterface is the interface for a Dec-POMDP with factored states. More...

#include <FactoredDecPOMDPDiscreteInterface.h>

Inheritance diagram for FactoredDecPOMDPDiscreteInterface:

Public Member Functions
virtual FactoredDecPOMDPDiscreteInterface *	Clone () const =0
	Returns a pointer to a copy of this class. More...

virtual const Scope &	GetAgentScopeForLRF (Index LRF) const =0
	Get all the agents in the scope of an LRF. More...

virtual const FactoredQFunctionScopeForStage	GetImmediateRewardScopes () const =0
	Returns all scopes of the immediate reward function in one object. More...

virtual RewardModel *	GetLRF (Index LRF) const =0
	Returns a pointer to the LRF-th reward model component. More...

virtual double	GetLRFReward (Index LRF, Index s_e, Index ja_e) const =0
	Returns reward for LRF, given ja_e - the joint index for a group action for the subset of agents as specified by the agent scope of LRF. More...

virtual double	GetLRFReward (Index LRF, const std::vector< Index > &sI_e, const std::vector< Index > &jaI_e) const =0

virtual double	GetLRFRewardFlat (Index LRF, Index flat_s, Index full_ja) const =0
	might be necessary? Returns reward for LRF, given a flat state index, and a full joint action. More...

virtual size_t	GetNrLRFs () const =0
	Get the number of LRFs in the factored representation. More...

virtual double	GetReward (const std::vector< Index > &sIs, const std::vector< Index > &aIs) const =0

virtual const Scope &	GetStateFactorScopeForLRF (Index LRF) const =0
	Get all the state factors in the scope of an LRF. More...

virtual void	Print () const =0

virtual Index	RestrictedActionVectorToJointIndex (Index LRF, const std::vector< Index > &actionVec_e) const =0
	convert an action vector of restricted scope to a joint index a_e. More...

virtual Index	RestrictedStateVectorToJointIndex (Index LRF, const std::vector< Index > &stateVec_e) const =0
	Get the vector of Factor indices corresponding to stateI. More...

virtual std::string	SoftPrint () const =0
	Prints some information on the MultiAgentDecisionProcessDiscreteInterface. More...

virtual	~FactoredDecPOMDPDiscreteInterface ()
	(default) Constructor More...

Public Member Functions inherited from DecPOMDPDiscreteInterface
virtual void	CreateNewRewardModel ()=0
	Creates a new reward model mapping. More...

virtual double	GetReward (Index sI, Index jaI) const =0
	Return the reward for state, joint action indices. More...

virtual RGet *	GetRGet () const =0

virtual void	SetReward (Index sI, Index jaI, double r)=0
	Set the reward for state, joint action indices. More...

virtual void	SetReward (Index sI, Index jaI, Index sucSI, double r)=0
	Set the reward for state, joint action , suc. state indices. More...

virtual void	SetReward (Index sI, Index jaI, Index sucSI, Index joI, double r)=0
	Set the reward for state, joint action, suc.state, joint obs indices. More...

virtual	~DecPOMDPDiscreteInterface ()
	import the GetReward function from the base class in current scope. More...

Public Member Functions inherited from DecPOMDPInterface
virtual double	GetDiscount () const =0
	Returns the discount parameter. More...

virtual double	GetReward (State s, JointAction ja) const =0
	Function that returns the reward for a state and joint action. More...

virtual reward_t	GetRewardType () const =0
	Returns the reward type. More...

virtual void	SetDiscount (double d)=0
	Sets the discount parameter to 0 < d <= 1. More...

virtual void	SetReward (State s, JointAction ja, double r)=0
	Function that sets the reward for a state and joint action. More...

virtual void	SetRewardType (reward_t r)=0
	Sets the reward type to reward_t r. More...

virtual	~DecPOMDPInterface ()
	Virtual destructor. More...

Public Member Functions inherited from POSGInterface
virtual double	GetDiscountForAgent (Index agentI) const =0
	Returns the discount parameter. More...

virtual double	GetRewardForAgent (Index agentI, State s, JointAction ja) const =0
	Function that returns the reward for a state and joint action. More...

virtual reward_t	GetRewardTypeForAgent (Index agentI) const =0
	Returns the reward type. More...

virtual void	SetDiscountForAgent (Index agentI, double d)=0
	Sets the discount parameter to 0 < d <= 1. More...

virtual void	SetRewardForAgent (Index agentI, State s, JointAction ja, double r)=0
	Function that sets the reward for an agent, state and joint action. More...

virtual void	SetRewardTypeForAgent (Index agentI, reward_t r)=0
	Sets the reward type to reward_t r. More...

virtual	~POSGInterface ()
	Virtual destructor. More...

Public Member Functions inherited from MultiAgentDecisionProcessInterface
virtual const Scope &	GetAllAgentScope () const =0

virtual size_t	GetNrAgents () const =0
	Return the number of agents. More...

virtual std::string	GetUnixName () const =0
	Returns the base part of the problem filename. More...

virtual	~MultiAgentDecisionProcessInterface ()
	Destructor. More...

Public Member Functions inherited from MultiAgentDecisionProcessDiscreteInterface
virtual const Action *	GetAction (Index agentI, Index a) const =0
	Return a ref to the a-th action of agent agentI. More...

virtual bool	GetEventObservability () const
	Whether observation models are P(o\|s,a) or P(o\|s,a,s'). More...

virtual double	GetInitialStateProbability (Index sI) const =0

virtual StateDistribution *	GetISD () const =0
	Returns the complete initial state distribution. More...

virtual const JointAction *	GetJointAction (Index i) const =0
	Return a ref to the i-th joint action. More...

virtual const JointObservation *	GetJointObservation (Index i) const =0
	Return a ref to the i-th joint observation. More...

virtual const std::vector < size_t > &	GetNrActions () const =0
	Return the number of actions vector. More...

virtual size_t	GetNrActions (Index AgentI) const =0
	Return the number of actions of agent agentI. More...

virtual size_t	GetNrJointActions () const =0
	Return the number of joint actions. More...

virtual size_t	GetNrJointActions (const Scope &agScope) const =0
	Get the number of joint actions the agents in agScope can form. More...

virtual size_t	GetNrJointObservations () const =0
	Return the number of joint observations. More...

virtual const std::vector < size_t > &	GetNrObservations () const =0
	Return the number of observations vector. More...

virtual size_t	GetNrObservations (Index AgentI) const =0
	Return the number of observations of agent agentI. More...

virtual size_t	GetNrStates () const =0
	Return the number of states. More...

virtual const Observation *	GetObservation (Index agentI, Index a) const =0
	Return a ref to the a-th observation of agent agentI. More...

virtual const ObservationModelDiscrete *	GetObservationModelDiscretePtr () const =0
	Returns a pointer to the underlying observation model. More...

virtual double	GetObservationProbability (Index jaI, Index sucSI, Index joI) const =0
	Return the probability of joint observation joI: P(joI\|jaI,sucSI). More...

virtual double	GetObservationProbability (Index sI, Index jaI, Index sucSI, Index joI) const

virtual OGet *	GetOGet () const =0

virtual const State *	GetState (Index i) const =0
	Returns a pointer to state i. More...

virtual TGet *	GetTGet () const =0

virtual const TransitionModelDiscrete *	GetTransitionModelDiscretePtr () const =0
	Returns a pointer to the underlying transition model. More...

virtual double	GetTransitionProbability (Index sI, Index jaI, Index sucSI) const =0
	Return the probability of successor state sucSI: P(sucSI\|sI,jaI). More...

virtual Index	IndividualToJointActionIndices (const Index *AI_ar) const =0
	Returns the joint action index that corresponds to the array of specified individual action indices. More...

virtual Index	IndividualToJointActionIndices (const std::vector< Index > &indivActionIndices) const =0
	Returns the joint action index that corresponds to the vector of specified individual action indices. More...

virtual Index	IndividualToJointActionIndices (const std::vector< Index > &ja_e, const Scope &agSC) const =0
	indiv->joint for a restricted set (Scope) of agents More...

virtual Index	IndividualToJointObservationIndices (const std::vector< Index > &indivObservationIndices) const =0
	Returns the joint observation index that corresponds to the vector of specified individual observation indices. More...

virtual Index	IndividualToJointObservationIndices (const std::vector< Index > &jo_e, const Scope &agSC) const
	indiv->joint for a restricted set (Scope) of agents More...

virtual const std::vector < Index > &	JointToIndividualActionIndices (Index jaI) const =0
	Returns a vector of indices to indiv. More...

virtual std::vector< Index >	JointToIndividualActionIndices (Index ja_e, const Scope &agSC) const =0
	joint->indiv for a restricted set (Scope) of agents More...

virtual const std::vector < Index > &	JointToIndividualObservationIndices (Index joI) const =0
	Returns a vector of indices to indiv. More...

virtual std::vector< Index >	JointToIndividualObservationIndices (Index jo_e, const Scope &agSC) const
	joint->indiv for a restricted set (Scope) of agents More...

virtual Index	JointToRestrictedJointActionIndex (Index jaI, const Scope &agSc_e) const =0
	Converts a global joint action index jaI to a restricted joint action index ja_e, for agents scope agSc_e Returns a vector of indices to indiv. More...

virtual Index	JointToRestrictedJointObservationIndex (Index joI, const Scope &agSc_e) const
	Converts a global joint observation index joI to a restricted joint observation index jo_e, for agents scope agSc_e Returns a vector of indices to indiv. More...

virtual Index	SampleInitialState () const =0
	Sample a state according to the initial state PDF. More...

virtual Index	SampleJointObservation (Index jaI, Index sucI) const =0
	Sample an observation - needed for simulations. More...

virtual Index	SampleJointObservation (Index sI, Index jaI, Index sucI) const

virtual Index	SampleSuccessorState (Index sI, Index jaI) const =0
	Sample a successor state - needed by simulations. More...

virtual	~MultiAgentDecisionProcessDiscreteInterface ()
	Destructor. More...

Public Member Functions inherited from MultiAgentDecisionProcessDiscreteFactoredStatesInterface
virtual Scope	AgentScopeBackup (const Scope &stateScope, const Scope &agentScope) const =0

virtual void	CacheFlatModels (bool sparse)

virtual Index	FactorValueIndicesToStateIndex (const std::vector< Index > &fv) const =0
	convert std::vector of (indices of) factor values to (flat) state index. More...

virtual Index	FactorValueIndicesToStateIndex (const std::vector< Index > &s_e_vec, const Scope &sfSC) const =0
	convert an local state vector s_e_vec of scope sfScope to a joint index. More...

virtual const TwoStageDynamicBayesianNetwork *	Get2DBN () const
	Copy assignment operator. More...

virtual const Scope &	GetAllStateFactorScope () const =0
	Convenience function to quickly get the full state scope. More...

virtual const FactoredStateDistribution *	GetFactoredISD () const =0

virtual size_t	GetNrStateFactorInstantiations (const Scope &sfScope) const =0
	Get the number of joint instantiations for the factors in sfScope. More...

virtual size_t	GetNrStateFactors () const =0
	Get the number of state components. -no is inherited from. More...

virtual const size_t	GetNrValuesForFactor (Index sf) const =0
	Get the number of possible values for a particular factor. More...

virtual const std::vector < size_t > &	GetNrValuesPerFactor () const =0
	Get the number of possible assignments or values to each factor. More...

virtual const StateFactorDiscrete *	GetStateFactorDiscrete (Index sfacI) const =0

virtual void	SampleInitialState (std::vector< Index > &sIs) const =0

virtual void	SampleJointObservation (const std::vector< Index > &aIs, const std::vector< Index > &sucIs, std::vector< Index > &oIs) const =0

virtual void	SampleSuccessorState (const std::vector< Index > &sIs, const std::vector< Index > &aIs, std::vector< Index > &sucIs) const =0

std::string	SoftPrintState (Index sI) const
	SoftPrints information on the MultiAgentDecisionProcessDiscrete. More...

virtual Index	StateIndexToFactorValueIndex (Index factor, Index s) const =0
	Get the value of a particular state factor given a joint flat state. More...

virtual std::vector< Index >	StateIndexToFactorValueIndices (Index stateI) const =0
	Get the vector of FactorValue indices corresponding to stateI used to be called virtual vector<Index> GetStateFactorValues(Index stateI) const. More...

virtual std::vector< Index >	StateIndexToFactorValueIndices (Index s_e, const Scope &sfSC) const =0
	convert an local state index s_e to a vector of state factors (with scope sfScope). More...

virtual Scope	StateScopeBackup (const Scope &stateScope, const Scope &agentScope) const =0

virtual	~MultiAgentDecisionProcessDiscreteFactoredStatesInterface ()
	(default) Constructor More...

Detailed Description

FactoredDecPOMDPDiscreteInterface is the interface for a Dec-POMDP with factored states.

It declares the get/set reward functions.

FactoredDecPOMDPDiscrete represents a factored DecPOMDPDiscrete FactoredDecPOMDPDiscreteInterface is an interface (i.e. pure abstract class) for a discrete factored DEC-POMDP model. Classes that implement this interface are, for instance, FactoredDecPOMDPDiscrete.

Because a factored Dec-POMDP can also be accessed as a regular Dec-POMDP (i.e., by indexing 'flat' states), this interface also derives from DecPOMDPDiscreteInterface (which defines the get/set-reward functions for a regular (non-factored) discrete Dec-POMDP).

The `LRFs' in this code refer to edges of the graph constructed from the factored immediate reward function. (i.e., each `edge' corresponds to a local reward function)

a `flat state' - a state s=<x_1...x_nrFactors> (which specifies a value for each state factor).

a `full joint action' - a joint action that specifies an action for each individual agent.

a `joint/group action' - an action for a subset of agents.

Constructor & Destructor Documentation

virtual FactoredDecPOMDPDiscreteInterface::~FactoredDecPOMDPDiscreteInterface ( )

inlinevirtual

(default) Constructor

Destructor.Can't make a virt.destr. pure abstract!

Member Function Documentation

virtual FactoredDecPOMDPDiscreteInterface* FactoredDecPOMDPDiscreteInterface::Clone ( ) const

pure virtual

Returns a pointer to a copy of this class.

Implements DecPOMDPDiscreteInterface.

Implemented in FactoredDecPOMDPDiscrete, ProblemFireFightingFactored, ProblemAloha, ProblemFOBSFireFightingFactored, and FactoredMMDPDiscrete.

virtual const Scope& FactoredDecPOMDPDiscreteInterface::GetAgentScopeForLRF ( Index LRF ) const

pure virtual

Get all the agents in the scope of an LRF.

Implemented in FactoredDecPOMDPDiscrete.

virtual const FactoredQFunctionScopeForStage FactoredDecPOMDPDiscreteInterface::GetImmediateRewardScopes ( ) const

pure virtual

Returns all scopes of the immediate reward function in one object.

Implemented in FactoredDecPOMDPDiscrete.

virtual RewardModel* FactoredDecPOMDPDiscreteInterface::GetLRF ( Index LRF ) const

pure virtual

Returns a pointer to the LRF-th reward model component.

Implemented in FactoredDecPOMDPDiscrete.

virtual double FactoredDecPOMDPDiscreteInterface::GetLRFReward	(	Index	LRF,
		Index	s_e,
		Index	ja_e
	)		const

pure virtual

Returns reward for LRF, given ja_e - the joint index for a group action for the subset of agents as specified by the agent scope of LRF.

s_e - the (joint) index of the subset of factors specified by the state factor scope of LRF.

For instance, let the agents scope of LRF be <1,3>, then group action <3,5> means that agent 1 select action 3, while agent 3 performs its 5th action. Using indextools we can find agSc_jaI. E.g. agSc_jaI = IndividualToJointIndices( <3,5>, <6,6> ) (where <6,6> is a vector which specifies the number of actions per agent, see IndexTools.h for more info).

Implemented in FactoredDecPOMDPDiscrete.

Referenced by FactoredQLastTimeStepOrElse::GetLocalQValue().

virtual double FactoredDecPOMDPDiscreteInterface::GetLRFReward	(	Index	LRF,
		const std::vector< Index > &	sI_e,
		const std::vector< Index > &	jaI_e
	)		const

pure virtual

Implemented in FactoredDecPOMDPDiscrete.

virtual double FactoredDecPOMDPDiscreteInterface::GetLRFRewardFlat	(	Index	LRF,
		Index	flat_s,
		Index	full_ja
	)		const

pure virtual

might be necessary? Returns reward for LRF, given a flat state index, and a full joint action.

Implemented in FactoredDecPOMDPDiscrete.

virtual size_t FactoredDecPOMDPDiscreteInterface::GetNrLRFs ( ) const

pure virtual

Get the number of LRFs in the factored representation.

Implemented in FactoredDecPOMDPDiscrete.

Referenced by FactoredQLastTimeStepOrElse::Initialize().

virtual double FactoredDecPOMDPDiscreteInterface::GetReward	(	const std::vector< Index > &	sIs,
		const std::vector< Index > &	aIs
	)		const

pure virtual

Implemented in FactoredDecPOMDPDiscrete.

Referenced by SimulationFactoredDecPOMDPDiscrete::Step().

virtual const Scope& FactoredDecPOMDPDiscreteInterface::GetStateFactorScopeForLRF ( Index LRF ) const

pure virtual

Get all the state factors in the scope of an LRF.

Implemented in FactoredDecPOMDPDiscrete.

Referenced by FactoredQLastTimeStepOrElse::GetStateFactorScopeForLQF().

virtual void FactoredDecPOMDPDiscreteInterface::Print ( ) const

pure virtual

Implemented in FactoredDecPOMDPDiscrete.

virtual Index FactoredDecPOMDPDiscreteInterface::RestrictedActionVectorToJointIndex	(	Index	LRF,
		const std::vector< Index > &	actionVec_e
	)		const

pure virtual

convert an action vector of restricted scope to a joint index a_e.

This is a convenience function that performs indiv->joint action index conversion for a specific edge e (LRF). (i.e., this function is typically called when requesting the immediate reward)

actionVec_e is an assignment of all actions in the agent scope scope of e.
the funtion returns a joint (group) index a_e.

Implemented in FactoredDecPOMDPDiscrete.

virtual Index FactoredDecPOMDPDiscreteInterface::RestrictedStateVectorToJointIndex	(	Index	LRF,
		const std::vector< Index > &	stateVec_e
	)		const

pure virtual

Get the vector of Factor indices corresponding to stateI.

convert a state vector of restricted scope to a joint index s_e.

This is a convenience function that performs indiv->joint state index conversion for a specific edge e (LRF).

stateVec_e is an assignment of all state factors in the state factor scope of e.
the funtion returns a joint (group) index s_e.

Implemented in FactoredDecPOMDPDiscrete.

Referenced by FactoredQLastTimeStepOrElse::GetLocalQValue().

virtual std::string FactoredDecPOMDPDiscreteInterface::SoftPrint ( ) const

pure virtual

Prints some information on the MultiAgentDecisionProcessDiscreteInterface.

Implements MultiAgentDecisionProcessDiscreteInterface.

Implemented in FactoredDecPOMDPDiscrete, and FactoredMMDPDiscrete.

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation