burlap.mdp.singleagent.model

Class FullModel.Helper

java.lang.Object

burlap.mdp.singleagent.model.FullModel.Helper

Enclosing interface:

FullModel

public static class FullModel.Helper
extends java.lang.Object

Constructor Summary

Constructors

Constructor and Description
Helper()

Method Summary

Modifier and Type	Method and Description
static java.util.List<TransitionProb>	deterministicTransition(SampleModel model, State s, Action a) Method to easily implement the FullModel.transitions(State, Action) method for deterministic domains.
static EnvironmentOutcome	sampleByEnumeration(FullModel model, State s, Action a) Method to implement the SampleModel.sample(State, Action) method when the FullModel.transitions(State, Action) method is implemented.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Helper

public Helper()

Method Detail

deterministicTransition

public static java.util.List<TransitionProb> deterministicTransition(SampleModel model,
                                                                     State s,
                                                                     Action a)

Method to easily implement the FullModel.transitions(State, Action) method for deterministic domains. Operates by getting a transition from the SampleModel.sample(State, Action) method and wraps it in a TransitionProb with probability 1 and then returns a list of that just one element.

Parameters:

model - the SampleModel with an implemented SampleModel.sample(State, Action) method.

s - the input state

a - the action taken.

Returns:

a List consisting of the single deterministic TransitionProb

sampleByEnumeration

public static EnvironmentOutcome sampleByEnumeration(FullModel model,
                                                     State s,
                                                     Action a)

Method to implement the SampleModel.sample(State, Action) method when the FullModel.transitions(State, Action) method is implemented. Operates by calling the FullModel.transitions(State, Action) method, rolls a random number, and selects a transition according the probability specified by FullModel.transitions(State, Action).

Parameters:

model - the FullModel with the implemented FullModel.transitions(State, Action) method.

s - the input state

a - the action to be applied in the input state

Returns:

a sampled transition (EnvironmentOutcome).