API tools faq
paste
Advertisement
mirosh111000

Q-learning(Maze)

mirosh111000
May 17th, 2024
95
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Python 19.09 KB | None | 0 0
raw download clone embed print report
  1. import numpy as np
  2. import seaborn as sns
  3. import matplotlib.pyplot as plt
  4.  
  5.  
  6. def create_maze():
  7.     maze = np.zeros((39, 39), dtype=float)
  8.     maze_idx = [i for i, j in np.ndenumerate(np.zeros((39, 39), dtype=int))]
  9.     wall = -4
  10.     maze[0, :] = wall
  11.     maze[-1, :] = wall
  12.     maze[:, 0] = wall
  13.     maze[:, -1] = wall
  14.  
  15.     idx = [
  16.         (4, 1), (8, 1), (12, 1), (18, 1), (20, 1), (28, 1), (34, 1),
  17.  
  18.         (2, 2), (4, 2), (6, 2), (8, 2), (10, 2), (12, 2), (14, 2), (16, 2), (17, 2), (18, 2), (20, 2), (22, 2),
  19.         (23, 2), (24, 2), (25, 2), (26, 2), (28, 2), (30, 2), (31, 2), (32, 2), (34, 2), (36, 2),
  20.  
  21.         (2, 3), (6, 3), (8, 3), (10, 3), (14, 3), (20, 3), (26, 3), (28, 3), (32, 3), (36, 3),
  22.  
  23.         (2, 4), (3, 4), (4, 4), (5, 4), (6, 4), (7, 4), (8, 4), (10, 4), (11, 4), (12, 4), (13, 4), (14, 4), (16, 4),
  24.         (18, 4), (19, 4), (20, 4), (22, 4), (23, 4), (24, 4), (26, 4), (27, 4,), (28, 4), (29, 4), (30, 4),
  25.         (32, 4), (33, 4), (34, 4), (35, 4), (36, 4),
  26.  
  27.         (4, 5), (8, 5), (10, 5), (12, 5), (16, 5), (22, 5), (24, 5), (26, 5), (32, 5), (36, 5),
  28.  
  29.         (1, 6), (2, 6), (4, 6), (6, 6), (8, 6), (10, 6), (12, 6), (14, 6), (15, 6), (16, 6), (17, 6), (18, 6), (19, 6),
  30.         (20, 6), (21, 6), (22, 6), (24, 6), (26, 6), (28, 6), (29, 6), (30, 6), (32, 6), (33, 6), (34, 6), (36, 6),
  31.  
  32.         (4, 7), (6, 7), (10, 7), (12, 7), (18, 7), (24, 7), (28, 7), (30, 7), (34, 7), (36, 7),
  33.  
  34.         (2, 8), (3, 8), (4, 8), (6, 8), (8, 8), (9, 8), (10, 8), (12, 8), (14, 8), (15, 8), (16, 8), (18, 8), (20, 8),
  35.         (21, 8), (22, 8), (23, 8), (24, 8), (25, 8), (26, 8), (27, 8), (28, 8), (30, 8), (32, 8), (34, 8), (36, 8),
  36.  
  37.         (2, 9), (6, 9), (8, 9), (12, 9), (16, 9), (24, 9), (28, 9), (32, 9), (36, 9),
  38.  
  39.         (2, 10), (4, 10), (5, 10), (6, 10), (8, 10), (10, 10), (11, 10), (12, 10), (13, 10), (14, 10), (16, 10), (17, 10),
  40.         (18, 10), (20, 10), (21, 10), (22, 10), (24, 10), (26, 10), (28, 10), (30, 10), (31, 10), (32, 10), (33, 10),
  41.         (34, 10), (36, 10),
  42.  
  43.         (2, 11), (4, 11), (8, 11), (12, 11), (16, 11), (22, 11), (24, 11), (26, 11), (28, 11), (30, 11), (34, 11), (36, 11),
  44.  
  45.         (2, 12), (3, 12), (4, 12), (6, 12), (7, 12), (8, 12), (9, 12), (10, 12), (12, 12), (14, 12), (15, 12), (16, 12),
  46.         (17, 12), (18, 12), (20, 12), (22, 12), (23, 12), (24, 12), (26, 12), (28, 12), (30, 12), (32, 12), (34, 12),
  47.         (36, 12), (37, 12),
  48.  
  49.         (6, 13), (12, 13), (18, 13), (20, 13), (24, 13), (26, 13), (30, 13), (32, 13),
  50.  
  51.         (2, 14), (4, 14), (6, 14), (8, 14), (9, 14), (10, 14), (12, 14), (13, 14), (14, 14), (15, 14), (16, 14), (18, 14),
  52.         (20, 14), (21, 14), (22, 14), (24, 14), (25, 14), (26, 14), (27, 14), (28, 14), (29, 14), (30, 14), (32, 14),
  53.         (33, 14), (34, 14), (35, 14), (36, 14),
  54.  
  55.         (2, 15), (4, 15), (6, 15), (8, 15), (10, 15), (16, 15), (18, 15), (20, 15), (22, 15), (24, 15), (30, 15), (32, 15),
  56.         (36, 15),
  57.  
  58.         (1, 16), (2, 16), (4, 16), (5, 16), (6, 16), (8, 16), (10, 16), (11, 16), (12, 16), (13, 16), (14, 16), (16, 16),
  59.         (17, 16), (18, 16), (20, 16), (22, 16), (24, 16), (26, 16), (27, 16), (28, 16), (30, 16), (31, 16), (32, 16),
  60.         (34, 16), (35, 16), (36, 16),
  61.  
  62.         (4, 17), (14, 17), (18, 17), (22, 17), (26, 17), (32, 17),
  63.  
  64.         (2, 18), (3, 18), (4, 18), (6, 18), (7, 18), (8, 18), (9, 18), (10, 18), (11, 18), (12, 18), (14, 18), (16, 18),
  65.         (18, 18), (19, 18), (20, 18), (21, 18), (22, 18), (23, 18), (24, 18), (25, 18), (26, 18), (27, 18), (28, 18),
  66.         (29, 18), (30, 18), (32, 18), (33, 18), (34, 18), (36, 18),
  67.  
  68.         (2, 19), (6, 19), (8, 19), (10, 19), (14, 19), (16, 19), (22, 19), (30, 19), (34, 19), (36, 19),
  69.  
  70.         (2, 20), (4, 20), (5, 20), (6, 20), (8, 20), (10, 20), (11, 20), (12, 20), (13, 20), (14, 20), (16, 20), (17, 20),
  71.         (18, 20), (19, 20), (20, 20), (22, 20), (24, 20), (25, 20), (26, 20), (27, 20), (28, 20), (30, 20), (32, 20),
  72.         (34, 20), (36, 20), (37, 20),
  73.  
  74.         (2, 21), (4, 21), (8, 21), (12, 21), (16, 21), (20, 21), (22, 21), (28, 21), (32, 21), (34, 21),
  75.  
  76.         (2, 22), (4, 22), (5, 22), (6, 22), (8, 22), (9, 22), (10, 22), (12, 22), (13, 22), (14, 22), (16, 22), (18, 22),
  77.         (20, 22), (22, 22), (23, 22), (24, 22), (25, 22), (26, 22), (28, 22), (29, 22), (30, 22), (31, 22), (32, 22),
  78.         (34, 22), (35, 22), (36, 22),
  79.  
  80.         (2, 23), (6, 23), (10, 23), (14, 23), (18, 23), (20, 23), (24, 23), (30, 23), (34, 23),
  81.  
  82.         (2, 24), (3, 24), (4, 24), (6, 24), (7, 24), (8, 24), (10, 24), (11, 24), (12, 24), (14, 24), (15, 24), (16, 24),
  83.         (17, 24), (18, 24), (20, 24), (21, 24), (22, 24), (24, 24), (25, 24), (26, 24), (27, 24), (28, 24), (29, 24),
  84.         (30, 24), (32, 24), (33, 24), (34, 24), (36, 24), (37, 24),
  85.  
  86.         (4, 25), (12, 25), (18, 25), (20, 25), (22, 25), (28, 25), (32, 25), (34, 25),
  87.  
  88.         (1, 26), (2, 26), (4, 26), (5, 26), (6, 26), (8, 26), (9, 26), (10, 26), (11, 26), (12, 26), (13, 26), (14, 26),
  89.         (15, 26), (16, 26), (18, 26), (20, 26), (22, 26), (23, 26), (24, 26), (26, 26), (27, 26), (28, 26), (30, 26),
  90.         (31, 26), (32, 26), (34, 26), (35, 26), (36, 26),
  91.  
  92.         (4, 27), (6, 27), (8, 27), (12, 27), (14, 27), (18, 27), (24, 27), (28, 27), (30, 27), (34, 27),
  93.  
  94.         (2, 28), (3, 28), (4, 28), (6, 28), (8, 28), (10, 28), (12, 28), (14, 28), (16, 28), (17, 28), (18, 28), (19, 28),
  95.         (20, 28), (21, 28), (22, 28), (23, 28), (24, 28), (25, 28), (26, 28), (28, 28), (30, 28), (32, 28), (33, 28),
  96.         (34, 28), (36, 28), (37, 28),
  97.  
  98.         (4, 29), (6, 29), (10, 29), (14, 29), (18, 29), (20, 29), (30, 29), (36, 29),
  99.  
  100.         (2, 30), (4, 30), (6, 30), (8, 30), (10, 30), (11, 30), (12, 30), (13, 30), (14, 30), (15, 30), (16, 30), (18, 30),
  101.         (20, 30), (22, 30), (23, 30), (24, 30), (25, 30), (26, 30), (27, 30), (28, 30), (29, 30), (30, 30), (31, 30),
  102.         (32, 30), (34, 30), (35, 30), (36, 30),
  103.  
  104.         (2, 31), (4, 31), (6, 31), (8, 31), (12, 31), (16, 31), (18, 31), (22, 31), (26, 31), (30, 31), (34, 31),
  105.  
  106.         (1, 32), (2, 32), (4, 32), (6, 32), (7, 32), (8, 32), (9, 32), (10, 32), (12, 32), (14, 32), (15, 32), (16, 32),
  107.         (18, 32), (19, 32), (20, 32), (21, 32), (22, 32), (24, 32), (26, 32), (28, 32), (29, 32), (30, 32), (32, 32),
  108.         (33, 32), (34, 32), (36, 32), (37, 32),
  109.  
  110.         (4, 33), (10, 33), (12, 33), (18, 33), (24, 33), (28, 33), (32, 33),
  111.  
  112.         (2, 34), (3, 34), (4, 34), (5, 34), (6, 34), (7, 34), (8, 34), (10, 34), (12, 34), (14, 34), (15, 34), (16, 34),
  113.         (18, 34), (20, 34), (21, 34), (22, 34), (23, 34), (24, 34), (25, 34), (26, 34), (28, 34), (30, 34), (31, 34),
  114.         (32, 34), (33, 34), (34, 34), (35, 34), (36, 34),
  115.  
  116.         (4, 35), (8, 35), (10, 35), (16, 35), (22, 35), (26, 35), (28, 35), (34, 35),
  117.  
  118.         (2, 36), (4, 36), (6, 36), (8, 36), (10, 36), (11, 36), (12, 36), (14, 36), (16, 36), (17, 36), (18, 36),
  119.         (20, 36), (22, 36), (24, 36), (26, 36), (27, 36), (28, 36), (29, 36), (30, 36), (32, 36), (34, 36), (36, 36),
  120.  
  121.         (2, 37), (6, 37), (8, 37), (14, 37), (18, 37), (20, 37), (24, 37), (30, 37), (32, 37), (36, 37),
  122.     ]
  123.     for i in idx: maze[i] = wall
  124.  
  125.     # Додавання виходу
  126.     exit = [(19, 38), (31, 38)]
  127.     maze[exit[0]] = -2
  128.     maze[exit[1]] = -2
  129.  
  130.     # Додавання стартової позиції
  131.     start_pos = (19, 0)
  132.     maze[start_pos] = -1
  133.  
  134.  
  135.     idx2 = [
  136.         (19, 1), (19, 2), (19, 3), (18, 3), (17, 3), (17, 4), (17, 5), (18, 5), (19, 5), (20, 5), (21, 5), (21, 4),
  137.         (21, 3), (22, 3), (23, 3), (24, 3), (25, 3), (25, 4), (25, 5), (25, 6), (25, 7), (26, 7), (27, 7), (27, 6),
  138.         (27, 5), (28, 5), (29, 5), (30, 5), (31, 5), (31, 6), (31, 7), (32, 7), (33, 7), (33, 8), (33, 9), (34, 9),
  139.         (35, 9), (35, 10), (35, 11), (35, 12), (35, 13), (36, 13), (37, 13), (37, 14), (37, 15), (37, 16), (37, 17),
  140.         (36, 17), (35, 17), (35, 18), (35, 19), (35, 20), (35, 21), (36, 21), (37, 21), (37, 22), (37, 23), (36, 23),
  141.         (35, 23), (35, 24), (35, 25), (36, 25), (37, 25), (37, 26), (37, 27), (36, 27), (35, 27), (35, 28), (35, 29),
  142.         (34, 29), (33, 29), (33, 30), (33, 31), (32, 31), (31, 31), (31, 32), (31, 33), (30, 33), (29, 33), (29, 34),
  143.         (29, 35), (30, 35), (31, 35), (31, 36), (31, 37)
  144.         ]
  145.  
  146.     idx1 = [
  147.         (19, 1), (19, 2), (19, 3), (18, 3), (17, 3), (16, 3), (15, 3), (15, 4), (15, 5), (14, 5), (13, 5), (13, 6),
  148.         (13, 7), (14, 7), (15, 7), (16, 7), (17, 7), (17, 8), (17, 9), (18, 9), (19, 9), (19, 10), (19, 11), (20, 11),
  149.         (21, 11), (21, 12), (21, 13), (22, 13), (23, 13), (23, 14), (23, 15), (23, 16), (23, 17), (24, 17), (25, 17),
  150.         (25, 16), (25, 15), (26, 15), (27, 15), (28, 15), (29, 15), (29, 16), (29, 17), (30, 17), (31, 17), (31, 18),
  151.         (31, 19), (32, 19), (33, 19), (33, 20), (33, 21), (33, 22), (33, 23), (32, 23), (31, 23), (31, 24), (31, 25),
  152.         (30, 25), (29, 25), (29, 26), (29, 27), (29, 28), (29, 29), (28, 29), (27, 29), (27, 28), (27, 27), (26, 27),
  153.         (25, 27), (25, 26), (25, 25), (24, 25), (23, 25), (23, 24), (23, 23), (22, 23), (21, 23), (21, 22), (21, 21),
  154.         (21, 20), (21, 19), (20, 19), (19, 19), (18, 19), (17, 19), (17, 18), (17, 17), (16, 17), (15, 17), (15, 16),
  155.         (15, 15), (14, 15), (13, 15), (12, 15), (11, 15), (11, 14), (11, 13), (10, 13), (9, 13), (8, 13), (7, 13),
  156.         (7, 14), (7, 15), (7, 16), (7, 17), (6, 17), (5, 17), (5, 18), (5, 19), (4, 19), (3, 19), (3, 20), (3, 21),
  157.         (3, 22), (3, 23), (4, 23), (5, 23), (5, 24), (5, 25), (6, 25), (7, 25), (7, 26), (7, 27), (7, 28), (7, 29),
  158.         (8, 29), (9, 29), (9, 30), (9, 31), (10, 31), (11, 31), (11, 32), (11, 33), (11, 34), (11, 35), (12, 35),
  159.         (13, 35), (13, 34), (13, 33), (14, 33), (15, 33), (16, 33), (17, 33), (17, 34), (17, 35), (18, 35), (19, 35),
  160.         (19, 36), (19, 37)
  161.         ]
  162.  
  163.     maze2 = np.copy(maze)
  164.     for i in idx1: maze2[i] = -0.5
  165.     for i in idx2: maze2[i] = -0.5
  166.        
  167.    
  168.  
  169.  
  170.     return maze, maze_idx, maze2, start_pos, exit
  171.  
  172.  
  173. class Maze:
  174.     def __init__(self, maze, maze_idx, maze_with_path, current_position, exit):
  175.         self.maze = maze  # Ігрове поле: 0 - порожньо, -1 - старт, -2 - вихід, -4 - стіна
  176.         self.maze_idx = maze_idx
  177.         self.maze_with_path = maze_with_path
  178.         self.current_position = current_position
  179.         self.start = current_position
  180.         self.exit = exit
  181.  
  182.    
  183.    
  184.     def display_maze(self, maze, path=[], arrow_path=[], title='Лабіринт'):
  185.         plt.figure(figsize=(10, 10))
  186.         plt.title(title)
  187.         if len(path) > 0:
  188.             for move, arrow in zip(path, arrow_path):
  189.                 plt.text(move[1]+0.15, move[0]+0.75, arrow, color='green', fontsize=10)
  190.         sns.heatmap(maze, cmap='hot', cbar=False, square=True, linewidth=0.5, linecolor='black', annot=False, fmt='')
  191.         plt.show()
  192.        
  193.    
  194.     def available_moves(self, pos):
  195.        
  196.         moves = [(pos[0]-1, pos[1]), (pos[0]+1, pos[1]), (pos[0], pos[1]-1), (pos[0], pos[1]+1)]
  197.        
  198.         correct_moves = []
  199.         for move in moves:
  200.             if move in self.maze_idx:
  201.                 correct_moves.append(move)
  202.        
  203.         return correct_moves
  204.    
  205.     def make_move(self, move):
  206.        
  207.         direction = None
  208.         if move[0] - self.current_position[0] == -1:
  209.             direction = "↑"
  210.         elif move[1] - self.current_position[1] == 1:
  211.             direction = "→"
  212.         elif move[0] - self.current_position[0] == 1:
  213.             direction = "↓"
  214.         elif move[1] - self.current_position[1] == -1:
  215.             direction = "←"
  216.        
  217.         self.current_position = move
  218.  
  219.         return direction
  220.    
  221.     def check_win(self):
  222.         if self.current_position not in (self.exit):
  223.             return 0
  224.         else:
  225.             return 1
  226.        
  227.     def get_state(self):
  228.         return self.current_position
  229.    
  230.     def get_reward(self, move):
  231.        
  232.         if self.maze[move] == 0:
  233.             return -1
  234.         elif self.maze[move] == -4:
  235.             return -1000
  236.         elif self.maze[move] == -2:
  237.             return 0
  238.         elif self.maze[move] == -1:
  239.             return -1
  240.        
  241.    
  242.     def train(self, agent, epochs=10, use_low_eps=False):
  243.         agent.rows, agent.cols = len(self.maze[0]), len(self.maze)
  244.         arr = [round(0.1*i*epochs) for i in range(1, 11)]
  245.         len_moves = []
  246.         for epoch in range(1, epochs+1):
  247.             if epoch in arr:
  248.                 if use_low_eps == True:
  249.                     agent.epsilon -= agent.epsilon*0.175
  250.  
  251.             path, arrow_path = [], []
  252.             self.current_position = self.start  # Скидання ігрового поля
  253.             while True:
  254.  
  255.                 state = self.current_position
  256.                 path.append(state)
  257.                
  258.                 action = agent.choose_move(self)
  259.                 direction = self.make_move(action)
  260.                 arrow_path.append(direction)
  261.                 next_state = self.current_position
  262.                 reward = self.get_reward(action)
  263.  
  264.                 agent.update_Q_values(self, state, action, reward, next_state)
  265.  
  266.                 if self.check_win() == 1:
  267.                     len_moves.append(len(path))
  268.                     if epoch % round(epochs*0.1) == 0:
  269.                         self.display_maze(self.maze, path=path, arrow_path=arrow_path, title=fr'epoch {epoch} - {next_state}')
  270.                     break
  271.                    
  272.         return np.arange(1, epochs+1), len_moves
  273.    
  274.     def play(self, agent, len_sh_path=86):
  275.         agent.rows, agent.cols, agent.epsilon = len(self.maze[0]), len(self.maze), 0.0
  276.         path, arrow_path = [], []
  277.         self.current_position = self.start  # Скидання ігрового поля
  278.         while True:
  279.  
  280.             state = self.current_position
  281.             path.append(state)
  282.  
  283.             action = agent.choose_move(self)
  284.             direction = self.make_move(action)
  285.             arrow_path.append(direction)
  286.             next_state = self.current_position
  287.             reward = self.get_reward(action)
  288.             if len(path) > 999:
  289.                 self.display_maze(self.maze, path=path, arrow_path=arrow_path, title=fr'Вихід не знайдено(Кількість ходів - {len(path)})')
  290.                 break
  291.  
  292.             if self.check_win() == 1:
  293.                 self.display_maze(self.maze, path=path, arrow_path=arrow_path, title=fr'Кількість ходів - {len(path)}(min={len_sh_path})')
  294.                 break
  295.        
  296.        
  297.                
  298.  
  299. class QLearningAgent:
  300.     def __init__(self, learning_rate=10**-4, discount_factor=0.9, epsilon=0.5):
  301.         self.learning_rate = learning_rate
  302.         self.discount_factor = discount_factor
  303.         self.epsilon = epsilon
  304.         self.Q_values = {}
  305.         self.rows = None
  306.         self.cols = None
  307.  
  308.     def choose_move(self, game):
  309.         # Вибір дії для наступного кроку
  310.         state = game.get_state()
  311.         move = game.available_moves(state)
  312.         if np.random.uniform(0, 1) < self.epsilon:
  313.             return move[np.random.randint(0, len(move))]
  314.         else:
  315.             # Вибір дії згідно з Q-таблицею
  316.             if state not in self.Q_values:
  317.                 # Якщо стан не існує у Q-таблиці, повертаємо випадкову дію
  318.                 return move[np.random.randint(0, len(move))]
  319.             else:
  320.                 # Вибираємо найкращу дію, враховуючи Q-значення
  321.                 q_values = self.Q_values[state]
  322.                 max_q_value = np.max(q_values)
  323.                 zero_indices = np.where(q_values == max_q_value)
  324.                 best_moves = [(i, j) for i, j in zip(zero_indices[0], zero_indices[1])]
  325.                 return best_moves[np.random.randint(0, len(best_moves))]
  326.  
  327.     def update_Q_values(self, game, state, action, reward, next_state):
  328.         # Оновлення значень Q-таблиці за допомогою формули Q-learning
  329.         if state not in self.Q_values:
  330.             maze_with_inf = np.full((self.rows, self.cols), -np.inf)
  331.             for i in game.available_moves(state):
  332.                 maze_with_inf[i] = 0.0
  333.             self.Q_values[state] = maze_with_inf
  334.            
  335.         if next_state not in self.Q_values:
  336.             maze_with_inf = np.full((self.rows, self.cols), -np.inf)
  337.             for i in game.available_moves(next_state):
  338.                 maze_with_inf[i] = 0.0
  339.             self.Q_values[next_state] = maze_with_inf
  340.  
  341.         # Обчислення нового Q-значення за формулою Q-learning
  342.         old_q_value = self.Q_values[state][action]
  343.         max_next_q_value = np.max(self.Q_values[next_state]) if next_state in self.Q_values else 0
  344.         new_q_value = old_q_value + self.learning_rate * (reward + self.discount_factor * max_next_q_value - old_q_value)
  345.  
  346.         # Оновлення Q-значення в таблиці
  347.         self.Q_values[state][action] = new_q_value
  348.  
  349. def create_small_maze():
  350.     maze = np.zeros((9, 9), dtype=float)
  351.     maze_idx = [i for i, j in np.ndenumerate(np.zeros((9, 9), dtype=int))]
  352.     wall = -4
  353.     maze[0, :] = wall
  354.     maze[-1, :] = wall
  355.     maze[:, 0] = wall
  356.     maze[:, -1] = wall
  357.    
  358.     idx = [
  359.         (2, 2), (3, 2), (4, 4), (5, 6), (7, 6), (7, 2), (3, 5), (2, 6), (3, 7), (3, 6), (3, 4), (5, 2), (6, 2),
  360.         (6, 4), (7, 4), (2, 4)
  361.     ]
  362.    
  363.     for i in idx: maze[i] = wall
  364.  
  365.     # Додавання виходу
  366.     exit = [(5, 8)]
  367.     maze[exit[0]] = -2
  368.    
  369.     # Додавання стартової позиції
  370.     start_pos = (1, 0)
  371.     maze[start_pos] = -1
  372.    
  373.     maze2 = np.copy(maze)
  374.     return maze, maze_idx, maze2, start_pos, exit
  375.            
  376.            
  377. def display_results(arange, len_moves):
  378.     plt.figure(figsize=(15, 6))
  379.     plt.title(f'Залежність кількості ходів від кількості епізодів навчання')
  380.     plt.plot(arange, len_moves)
  381.     plt.xlabel(f'Номер епізоду')
  382.     plt.ylabel(f'Кількість ходів')
  383.     plt.grid()
  384.     plt.show()
  385.  
  386.    
  387. maze_small, maze_idx_small, maze_with_path_small, current_position_small, exit_small = create_small_maze()
  388.  
  389. game_small = Maze(maze_small, maze_idx_small, maze_with_path_small, current_position_small, exit_small)
  390. game_small.display_maze(game_small.maze)
  391.  
  392. agent_small = QLearningAgent(epsilon=0.2, learning_rate=10**-2)
  393. arange_small, len_moves_small = game_small.train(agent=agent_small, epochs=10000, use_low_eps=False)
  394. game_small.play(agent_small, len_sh_path=14)
  395. display_results(arange_small, len_moves_small)
  396.  
  397.  
  398. maze, maze_idx, maze_with_path, current_position, exit = create_maze()
  399.  
  400.  
  401. game = Maze(maze, maze_idx, maze_with_path, current_position, exit)
  402. game.display_maze(game.maze)
  403. game.display_maze(game.maze_with_path)
  404. agent = QLearningAgent(epsilon=0.2, learning_rate=10**-2)
  405. arange, len_moves = game.train(agent, epochs=25000, use_low_eps=True)
  406. game.play(agent)
  407. display_results(arange, len_moves)
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!