burlap.oomdp.singleagent

Class FullActionModel.FullActionModelHelper

java.lang.Object

burlap.oomdp.singleagent.FullActionModel.FullActionModelHelper

Enclosing interface:

FullActionModel

public static class FullActionModel.FullActionModelHelper
extends java.lang.Object

A class with helper methods for working with actions that implement FullActionModel.

Constructor Summary

Constructors

Constructor and Description
FullActionModel.FullActionModelHelper()

Method Summary

Methods

Modifier and Type	Method and Description
static java.util.List<TransitionProbability>	removeZeroProbTransitions(java.util.List<TransitionProbability> sourceTransitions) Takes a list of fully enumerated State transitions specified with TransitionProbability objects, and returns a new list of TransitionProbability objects that excludes any objects assigned probability 0.
static State	sampleTransitionFromEnumeratedDistribution(State previousState, GroundedAction action) Samples a state from fully enumerated transition dynamics that are defined with the FullActionModel FullActionModel.getTransitions(burlap.oomdp.core.states.State, GroundedAction) method.

Methods inherited from class java.lang.Object

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

Constructor Detail

FullActionModel.FullActionModelHelper

public FullActionModel.FullActionModelHelper()

Method Detail

sampleTransitionFromEnumeratedDistribution

public static State sampleTransitionFromEnumeratedDistribution(State previousState,
                                               GroundedAction action)

Samples a state from fully enumerated transition dynamics that are defined with the FullActionModel FullActionModel.getTransitions(burlap.oomdp.core.states.State, GroundedAction) method. This method is implement by first getting the enumerated transitions, rolling a random number, and selecting a next state transition from the enumerated transitions according to their assigned probability.

Parameters:

previousState - the State in which an action is to be performed

action - the GroundedAction to apply to previousState

Returns:

a sampled State

removeZeroProbTransitions

public static java.util.List<TransitionProbability> removeZeroProbTransitions(java.util.List<TransitionProbability> sourceTransitions)

Takes a list of fully enumerated State transitions specified with TransitionProbability objects, and returns a new list of TransitionProbability objects that excludes any objects assigned probability 0.

Parameters:

sourceTransitions - the original list of state transitions

Returns:

a new list of TransitionProbability objects without any with probability 0.