import numpy as np
import tkinter as tk
import copy
import pickle
from Q_Learning_Tic_Tac_Toe import Game, QPlayer, RandomPlayer # Classes used for Tic Tac Toe
root = tk.Tk()
epsilon = 2e-2
lr = 10e-3
player1 = QPlayer(mark="X", epsilon=epsilon,
learningRate=lr) ### add parameters!
player2 = RandomPlayer(mark="O")
game = Game(root, player1, player2) #,board_size=5,streak_size=4)
N_episodes = 1000
for episodes in range(N_episodes):
game.play()
game.reset()
Q = game.Q
filename = "Q_epsilon_09_Nepisodes_{}.p".format(N_episodes)
pickle.dump(Q, open(filename, "wb"))
ws1 = wb.active
ws1.title = "Training Results"
i = 1
from Q_Learning_Tic_Tac_Toe import Game, HumanPlayer, QPlayer, Board
board = Board()
Q = pickle.load(open("Q_epsilon_09_Nepisodes_20000.p", "rb"))
root = tk.Tk()
player1 = QPlayer(mark="X", epsilon=0.01)
player2 = QPlayer(mark="O", epsilon=0.01)
game = Game(root, player1, player2, Q=Q)
N_episodes = 20000
for episodes in range(N_episodes):
game.play()
game.reset()
if game.board.winner() is None:
ws1.cell(row=i + 1, column=1, value=0)
else:
if game.current_player.mark == "X":
ws1.cell(row=i + 1, column=1, value=1)
elif game.current_player.mark == "O":
ws1.cell(row=i + 1, column=1, value=-1)
i += 1
import numpy as np
import tkinter as tk
import copy
from tqdm import tqdm
import pickle
from Q_Learning_Tic_Tac_Toe import Game, QPlayer # Classes used for Tic Tac Toe
root = tk.Tk()
epsilon = 0.9
player1 = QPlayer(mark="X", epsilon=epsilon)
player2 = QPlayer(mark="O", epsilon=epsilon)
game = Game(root, player1, player2)
N_episodes = 200000
for episodes in tqdm(range(N_episodes)):
game.play()
game.reset(log=False)
Q = game.Q
filename = "Q_epsilon_09_Nepisodes_{}.p".format(N_episodes)
pickle.dump(Q, open(filename, "wb"))