burlap.behavior.singleagent.learnbydemo.mlirl.differentiableplanners.diffvinit

Class LinearStateDiffVF

java.lang.Object

burlap.behavior.singleagent.learnbydemo.mlirl.differentiableplanners.diffvinit.DifferentiableVInit.ParamedDiffVInit

burlap.behavior.singleagent.learnbydemo.mlirl.differentiableplanners.diffvinit.LinearStateDiffVF

All Implemented Interfaces:

DifferentiableVInit, ValueFunction, ValueFunctionInitialization

public class LinearStateDiffVF
extends DifferentiableVInit.ParamedDiffVInit

A class for defining a (differentiable) linear function over state features for value function initialization. This class is useful for learning the value function initialization for leaf nodes of a finite horizon planner with MLIRL.

Author:

James MacGlashan.

Nested Class Summary

Nested classes/interfaces inherited from interface burlap.behavior.singleagent.learnbydemo.mlirl.differentiableplanners.diffvinit.DifferentiableVInit

DifferentiableVInit.ParamedDiffVInit

Nested classes/interfaces inherited from interface burlap.behavior.singleagent.ValueFunctionInitialization

ValueFunctionInitialization.ConstantValueFunctionInitialization

Field Summary

Fields

Modifier and Type	Field and Description
protected StateToFeatureVectorGenerator	fvgen The state feature vector generator over which the linear function operates

Fields inherited from class burlap.behavior.singleagent.learnbydemo.mlirl.differentiableplanners.diffvinit.DifferentiableVInit.ParamedDiffVInit

dim, parameters

Constructor Summary

Constructors

Constructor and Description
LinearStateDiffVF(StateToFeatureVectorGenerator fvgen, int dim) Initializes with the state feature vector generator over which the linear function is defined and the dimensionality of it.

Method Summary

Methods

Modifier and Type	Method and Description
double[]	getQGradient(State s, AbstractGroundedAction ga) Returns the Q-value function gradient.
double[]	getVGradient(State s) Returns the value function gradient.
double	qValue(State s, AbstractGroundedAction a) Returns the initialization value of the Q-value function for a given state and action pair.
double	value(State s) Returns the value function evaluation of the given state.

Methods inherited from class burlap.behavior.singleagent.learnbydemo.mlirl.differentiableplanners.diffvinit.DifferentiableVInit.ParamedDiffVInit

getParameterDimension, getParameters, randomizeParameters, setParameter, setParameters

Methods inherited from class java.lang.Object

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

Field Detail

fvgen

protected StateToFeatureVectorGenerator fvgen

The state feature vector generator over which the linear function operates

Constructor Detail

LinearStateDiffVF

public LinearStateDiffVF(StateToFeatureVectorGenerator fvgen,
                 int dim)

Initializes with the state feature vector generator over which the linear function is defined and the dimensionality of it.

Parameters:

fvgen - the state feature vector generator over which the linear function is defined.

dim - the dimensionality of the feature vector/parameters

Method Detail

getVGradient

public double[] getVGradient(State s)

Description copied from interface: DifferentiableVInit

Returns the value function gradient.

Parameters:

s - the state on which the value function is to be evaluated

Returns:

the value function gradient.

getQGradient

public double[] getQGradient(State s,
                    AbstractGroundedAction ga)

Description copied from interface: DifferentiableVInit

Returns the Q-value function gradient.

Parameters:

s - the state on which the Q-value is to be evaluated.

ga - the action on which the Q-value is to be evaluated.

Returns:

the Q-value function gradient

value

public double value(State s)

Description copied from interface: ValueFunction

Returns the value function evaluation of the given state. If the value is not stored, then the default value specified by the ValueFunctionInitialization object of this class is returned.

Parameters:

s - the state to evaluate.

Returns:

the value function evaluation of the given state.

qValue

public double qValue(State s,
            AbstractGroundedAction a)

Description copied from interface: ValueFunctionInitialization

Returns the initialization value of the Q-value function for a given state and action pair.

Parameters:

s - the state for which to get the initial value of the Q-value function.

a - the action for which to get the initial value of the Q-value function.

Returns:

the initialization value of the Q-value function for a given state and action pair.