RLPark 1.0.0
Reinforcement Learning Framework in Java
|
00001 package rlpark.plugin.rltoys.envio.policy; 00002 00003 import java.util.Random; 00004 00005 import rlpark.plugin.rltoys.envio.actions.Action; 00006 import rlpark.plugin.rltoys.math.vector.RealVector; 00007 00008 public class ConstantPolicy extends StochasticPolicy { 00009 private static final long serialVersionUID = 9106677500699183729L; 00010 protected final double[] distribution; 00011 00012 public ConstantPolicy(Random random, Action[] actions, double[] distribution) { 00013 super(random, actions); 00014 assert actions.length == distribution.length; 00015 this.distribution = distribution; 00016 } 00017 00018 @Override 00019 public double pi(Action a) { 00020 return distribution[atoi(a)]; 00021 } 00022 00023 @Override 00024 public Action sampleAction() { 00025 return chooseAction(distribution); 00026 } 00027 00028 @Override 00029 public double[] distribution() { 00030 return distribution; 00031 } 00032 00033 @Override 00034 public void update(RealVector x) { 00035 } 00036 }