burlap.behavior.singleagent.learnfromdemo.mlirl.support

Class QGradientTuple

java.lang.Object

burlap.behavior.singleagent.learnfromdemo.mlirl.support.QGradientTuple

public class QGradientTuple
extends java.lang.Object

A tuple (triple) for storing the Q-gradient associated with a state and action.

Author:

James MacGlashan.

Field Summary

Fields

Modifier and Type	Field and Description
AbstractGroundedAction	a The action
FunctionGradient	gradient The gradient for the state and action.
State	s The state

Constructor Summary

Constructors

Constructor and Description
QGradientTuple(State s, AbstractGroundedAction a, FunctionGradient gradient) Initializes.

Method Summary

Methods inherited from class java.lang.Object

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

Field Detail

s

public State s

The state

a

public AbstractGroundedAction a

The action

gradient

public FunctionGradient gradient

The gradient for the state and action.

Constructor Detail

QGradientTuple

public QGradientTuple(State s,
              AbstractGroundedAction a,
              FunctionGradient gradient)

Initializes.

Parameters:

s - the state

a - the action

gradient - the gradient for the state an action