public static void GetStateAndAction(int i, int j, int k, out InvertedPendulumState state, out InvertedPendulumAction action)
{
GetState(j, k, out state);
GetAction(i, out action);
}
public static void GetActionIndices(InvertedPendulumAction action, out int i)
{
i = action.action + 1;
}
public float[, ,] value = new float[3, LEN, LEN];//-1,0,1;-pi..pi;-10..10
public static void GetIndices(InvertedPendulumState state, InvertedPendulumAction action, out int i, out int j, out int k)
{
GetStateIndices(state, out j, out k);
GetActionIndices(action, out i);
}
public void GetBestActionAndUtilityForState(ReinforcementLearningState state, out ReinforcementLearningAction action, out float retval)
{
float max = float.MinValue;
int maxact = 0;
int i;
int j;
int k;
GetStateIndices((InvertedPendulumState)state, out j, out k);
for (int act = -1; act <= 1; ++act)
{
GetActionIndices(new InvertedPendulumAction(act), out i);
if (value[i, j, k] > max)
{
max = value[i, j, k];
maxact = act;
}
}
if (value[1, j, k] == max) maxact = 0;
action = new InvertedPendulumAction(maxact);
retval = max;
}
public static void GetAction(int i, out InvertedPendulumAction action)
{
action = new InvertedPendulumAction(i - 1);
}
private void button5_Click(object sender, EventArgs e)
{
adp.Train();
customControl21.Invalidate();
Random r = new Random();
InvertedPendulumQStore QStore = (InvertedPendulumQStore)adp.qstore;
for (int aa = 0; aa < 100000; ++aa)
{
InvertedPendulumState state = new InvertedPendulumState((float)(2 * r.NextDouble() - 1) * Math.PI, (2 * r.NextDouble() - 1) * 3);
InvertedPendulumAction action = new InvertedPendulumAction(r.Next(3) - 1);
((InvertedPendulumESIGMNQStore)esigmn.qFunction).igmn.Train(new Vector(new double[] { state.a, state.w, action.action, QStore.Evaluate(state, action) }));
}
//for (int j = 0; j < QStore.value.GetLength(1); ++j)
//{
// for (int k = 0; k < QStore.value.GetLength(2); ++k)
// {
// for (int i = 0; i < QStore.value.GetLength(0); ++i)
// {
// InvertedPendulumState state;
// InvertedPendulumAction action;
// InvertedPendulumQStore.GetStateAndAction(i, j, k, out state, out action);
// ((InvertedPendulumESIGMNQStore)esigmn.qFunction).igmn.Train(new Vector(new double[] { state.a, state.w, action.action, QStore.value[i, j, k] }));
// }
// }
//}
esigmnViewer1.Invalidate();
}
