private int[] Radar(Prostredie env)
{
int[] scan = new int[8];
// Hore
if (!JeMimoAreny(currentPos.x, currentPos.y - 1, 10, 10))
{
scan[0] = env.prostredie[currentPos.y - 1][currentPos.x].id;
}
// Vpravo-Hore
if (!JeMimoAreny(currentPos.x + 1, currentPos.y - 1, 10, 10))
{
scan[1] = env.prostredie[currentPos.y - 1][currentPos.x + 1].id;
}
// Vpravo
if (!JeMimoAreny(currentPos.x + 1, currentPos.y, 10, 10))
{
scan[2] = env.prostredie[currentPos.y][currentPos.x + 1].id;
}
// Vpravo-Dole
if (!JeMimoAreny(currentPos.x + 1, currentPos.y + 1, 10, 10))
{
scan[3] = env.prostredie[currentPos.y + 1][currentPos.x + 1].id;
}
// Dole
if (!JeMimoAreny(currentPos.x, currentPos.y + 1, 10, 10))
{
scan[4] = env.prostredie[currentPos.y + 1][currentPos.x].id;
}
// Vlavo-Dole
if (!JeMimoAreny(currentPos.x - 1, currentPos.y + 1, 10, 10))
{
scan[5] = env.prostredie[currentPos.y + 1][currentPos.x - 1].id;
}
// Vlavo
if (!JeMimoAreny(currentPos.x - 1, currentPos.y, 10, 10))
{
scan[6] = env.prostredie[currentPos.y][currentPos.x - 1].id;
}
// Vlavo-Hore
if (!JeMimoAreny(currentPos.x - 1, currentPos.y - 1, 10, 10))
{
scan[7] = env.prostredie[currentPos.y - 1][currentPos.x - 1].id;
}
return(scan);
}
public void reset(Prostredie env, int t = 0, bool training = true)
{
// Vymaz vsetky jablka a miny
env.NahradObjekty(Jablko.Tag, new Cesta());
env.NahradObjekty(Mina.Tag, new Cesta());
if (training == true)
{
var idx = r.Next(0, 3);
this.currentPos = new Vector2(Prostredie.startPositionX_training[idx], Prostredie.startPositionY_training[idx]);
apple_count = r.Next(2, 5) + 1;
for (int i = 0; i < apple_count; i++)
{
env.GenerateItem(new Jablko());
}
mine_count = r.Next(0, 3) + 1;
for (int i = 0; i < mine_count; i++)
{
env.GenerateItem(new Mina());
}
}
else
{
this.currentPos = new Vector2(this.testing_startsPos[t].x, this.testing_startsPos[t].y);
for (int i = 0; i < this.testing_applesPos[t].Length; i++)
{
env.prostredie[this.testing_applesPos[t][i].y][this.testing_applesPos[t][i].x] = new Jablko();
}
for (int i = 0; i < this.testing_minesPos[t].Length; i++)
{
env.prostredie[this.testing_minesPos[t][i].y][this.testing_minesPos[t][i].x] = new Mina();
}
}
stav = new AI.QLearning.QState
{
PositionX = currentPos.x,
PositionY = currentPos.y,
stateRadar = Radar(env)
};
this.apples = 0;
this.mines = 0;
}
public bool AktualizujAgenta(Prostredie env, bool ucenie, double eps, out float odmena)
{
bool isValid;
int akcia;
// Vyber akciu
var akcia_max = qBrain.NajdiMaxAkciu(stav);
/****************************************************/
/* Agent vykona akciu */
/****************************************************/
// Ak existuje vedomost
if (akcia_max != null)
{
// explore
if (ucenie && r.NextDouble() < eps)
{
isValid = sample(out akcia);
}
else
{
akcia = akcia_max.Value;
isValid = Pohyb((EAkcie)akcia, 10, 10);
}
}
// Ak neexistuje este vedomost
else
{
isValid = sample(out akcia);
// ... vytvor zaznam o najdenom stave
qBrain.Qtable.Add(stav, new float[] { 0f, 0f, 0f, 0f });
}
if (isValid)
{
odmena = env.Hodnotenie(currentPos.x, currentPos.y);
}
else
{
odmena = -1.0f;
}
/****************************************************/
/* Feedback */
/****************************************************/
var novyStav = new AI.QLearning.QState
{
PositionX = currentPos.x,
PositionY = currentPos.y,
stateRadar = Radar(env)
};
if (ucenie)
{
var buducaAkcia = qBrain.NajdiMaxAkciu(novyStav);
var buducaQhodnota = 0f;
// Aktualizuj Qtable hodnotu pre [s;a]
if (buducaAkcia != null)
{
buducaQhodnota = qBrain.Qtable[novyStav][buducaAkcia.Value];
}
qBrain.Aktualizuj(stav, akcia, odmena, buducaQhodnota);
}
this.stav = novyStav;
// Agent zobral jablko
if (env.prostredie[currentPos.y][currentPos.x].id == Jablko.Tag)
{
env.prostredie[currentPos.y][currentPos.x] = new Cesta();
this.apples += 1;
}
// Agent aktivoval minu
if (env.prostredie[currentPos.y][currentPos.x].id == Mina.Tag)
{
env.prostredie[currentPos.y][currentPos.x] = new Cesta();
this.mines += 1;
}
return(isValid);
}
private static Statistics run(Prostredie env, Agent a, int episodes, int steps, bool training = true)
{
Statistics stat = new Statistics();
var epsilon = training == true ? 1.0f : 0.0f;
float score;
int is_end;
// Trening agenta
for (int episode = 0, step; episode < episodes; episode++)
{
var watch = System.Diagnostics.Stopwatch.StartNew();
a.reset(env, episode, training);
is_end = 0;
score = 0;
for (step = 0; step < steps; step++)
{
if (training == false)
{
Console.Clear();
env.Vypis(a.currentPos.x, a.currentPos.y);
Thread.Sleep(200);
}
var isValid = a.AktualizujAgenta(env, training, epsilon, out float odmena);
score += odmena;
// ukonci hru ak nasiel ciel
if (env.prostredie[a.currentPos.y][a.currentPos.x].id == Vychod.Tag)
{
is_end = 100;
break;
}
else if (isValid == false)
{
break;
}
}
watch.Stop();
if (epsilon >= 0.01f)
{
epsilon *= epsilon_decay;
}
/*if ((episode % 1000) == 0)
* {
* Console.WriteLine($"\nepsilon: {epsilon}, epoch: {episode}/{episodes}");
* Console.WriteLine($"Pocet naucenych stavov: {a.PocetUlozenychStavov}\n");
* Console.WriteLine($"apples: {a.apples}/{a.apple_count}, mines: {a.mines}/{a.mine_count}");
* }*/
// log only testing phase
if (training == false)
{
var apple = (a.apples / (float)a.apple_count) * 100.0f;
var mine = (a.mines / (float)a.mine_count) * 100.0f;
log_file.WriteLine($"{episode};{score};{step};{watch.Elapsed.TotalMilliseconds * 1000};{apple};{mine};{is_end}");
//Console.WriteLine($"{apple};{mine};{is_end}");
stat.append(apple, mine, is_end);
}
}
return(stat);
}