public override double[] PerformAction(double[] action)
{
Debug.DLogHeader("SEASON TASK START", true);
Debug.DLog($"{"Action:",-16} {Utilities.ToString(action, "f4")}", true);
Debug.DLog($"{"Step:",-16} {_step}", true);
double thisScore = 0;
double[] observation = new double[OutputCount];
int task = _step % 2; // 0 = reward / food step, 1 = eat step
switch (task)
{
case 0:
Debug.DLog("Task: Reward / Food Step", true);
double eatVal = action[0];
thisScore = Evaluate(eatVal, (_step - 2) / 2); // we compare against the previos food, therefor - 2
observation = GetOutput(_step, thisScore);
if (!ScoreThisStep((_step - 2) / 2)) // no scoring here
{
thisScore = 0;
}
_score += thisScore;
Debug.DLog($"{"Eating:",-16} {eatVal}" +
$"\n{"Last Was Poisonous:",-16} {Sequence[(_step - 2) / 2].IsPoisonous}" +
$"\n{"Score:",-16} {thisScore.ToString("F4")}" +
$"\n{"Total Score:",-16} {_score.ToString("F4")} / {(_step / 2)}" +
$"\n{"Max Score:",-16} {Sequence.Length.ToString("F4")}", true);
break;
case 1:
Debug.DLog("Task: Eat Step", true);
// send only 0s
break;
default:
break;
}
Debug.DLogHeader("SEASON TASK END", true);
if (RecordTimeSteps)
{
_prevTimeStep = new EnvironmentTimeStep(action, observation, thisScore);
}
_step++;
return observation;
}
public override double[] PerformAction(double[] action)
{
Debug.DLogHeader("SEASON TASK START", true);
Debug.DLog($"{"Action:",-16} {Utilities.ToString(action, "f4")}", true);
Debug.DLog($"{"Step:",-16} {_step}", true);
double thisEvaluation = 0;
double[] observation;
int task = _step % StepNum;
if (task == StepNum - 1)
{
Debug.DLog($"Task: Reward Step", true);
double eatVal = action[0];
thisEvaluation = Evaluate(eatVal, (_step / StepNum));
observation = GetOutput(_step, thisEvaluation);
if (!ScoreThisStep((_step / StepNum)))
{
thisEvaluation = 0;
}
_score += thisEvaluation;
Debug.DLog($"{"Eating:",-16} {eatVal}" +
$"\n{"Poisonous:",-16} {Sequence[(_step / 3) - 1].IsPoisonous}" +
$"\n{"Score:",-16} {thisEvaluation.ToString("F4")}" +
$"\n{"Total Score:",-16} {_score.ToString("F4")} / {(_step / 3) - 1}" +
$"\n{"Max Score:",-16} {Sequence.Length.ToString("F4")}", true);
}
else
{
Debug.DLog($"Task: Normal Step {(_step / StepNum) - 1}", true);
observation = GetOutput(_step, -1);
}
Debug.DLogHeader("SEASON TASK END", true);
if (RecordTimeSteps)
{
_prevTimeStep = new EnvironmentTimeStep(action, observation, thisEvaluation);
}
_step++;
return observation;
}
public void Record(EnvironmentTimeStep environmentTimeStep)
{
_recordedTimeSteps.Add(new TimeStep(environmentTimeStep));
}
public void Record(EnvironmentTimeStep environmentTimeStep, TuringMachineTimeStep turingTimeStep)
{
_recordedTimeSteps.Add(new TimeStep(environmentTimeStep, turingTimeStep));
}
public TimeStep(EnvironmentTimeStep environmentTimeStep)
{
EnvironmentTimeStep = environmentTimeStep;
TuringMachineTimeStep = default(TuringMachineTimeStep);
}
public TimeStep(EnvironmentTimeStep environmentTimeStep, TuringMachineTimeStep turingMachineTimeStep)
{
EnvironmentTimeStep = environmentTimeStep;
TuringMachineTimeStep = turingMachineTimeStep;
}
public override double[] PerformAction(double[] action)
{
double[] prev = _sequence[_step - 1];
double target = prev[0] == 1d ^ prev[1] == 1d ? 1d : 0d;
double thisScore = Absolute(target, action[0]);
_score += thisScore;
double[] result = GetObservation(_step);
_step++;
if (RecordTimeSteps)
{
_prevTimeStep = new EnvironmentTimeStep(action, result, thisScore);
}
return result;
}
public TimeStep(EnvironmentTimeStep environmentTimeStep)
{
EnvironmentTimeStep = environmentTimeStep;
TuringMachineTimeStep = default(TuringMachineTimeStep);
}
public TimeStep(EnvironmentTimeStep environmentTimeStep, TuringMachineTimeStep turingMachineTimeStep)
{
EnvironmentTimeStep = environmentTimeStep;
TuringMachineTimeStep = turingMachineTimeStep;
}
public void Record(EnvironmentTimeStep environmentTimeStep)
{
_recordedTimeSteps.Add(new TimeStep(environmentTimeStep));
}
public void Record(EnvironmentTimeStep environmentTimeStep, TuringMachineTimeStep turingTimeStep)
{
_recordedTimeSteps.Add(new TimeStep(environmentTimeStep, turingTimeStep));
}