public Action GetLearningAction(QState state)
{
if (!ModelReady())
{
Iteration = 1;
Initialize(state.GridSize, state.VectorSize, state.Depth);
}
if (!_isFirstTurn)
{
if (state.IsTerminal)
{
StoreSARS(new SARS(_prevState, _prevAction, state));
_isFirstTurn = true;
return(null);
}
if (state.Equals(_prevState) && (Discretize || (!Discretize && _stateCounter <= 10)))
{
_stateCounter++;
return(_prevAction.Action);
}
StoreSARS(new SARS(_prevState, _prevAction, state));
}
var a = EpsilonGreedy(Epsilon(Iteration), state);
_prevAction = a;
_prevState = state;
_stateCounter = 0;
_isFirstTurn = false;
_trainingCounter++;
if (_trainingCounter >= TrainingInterval && Time.timeScale != 0)
{
var ts = Time.timeScale;
_trainingCounter = 0;
Time.timeScale = 0;
QAIManager.RunCoroutine(PrioritySweeping ? RunPriotizedTraining(ts) : RunTraining(ts));
}
return(a.Action);
}
public override bool Equals(object obj)
{
return(state.Equals(((QStateActionPair)obj).state) && action == ((QStateActionPair)obj).action);
}
public bool Equals(SARS other)
{
return(State.Equals(other.State) && NextState.Equals(other.NextState) && Action.Equals(other.Action));
}