ReinforcementLearning/QTableAttempt.py at main · loregi01/ReinforcementLearning · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
import gym
from gym import Env
import numpy as np
import random
from collections import defaultdict
import matplotlib.pyplot as plt

"""
For the visualization part:
  env = gym.make('MountainCar-v0', render_mode="human")
  env.render()
"""

def moving_average(a, n):
  ret = np.cumsum(a, dtype=float)
  ret[n:] = ret[n:] - ret[:-n]
  return ret[n - 1:] / n

learning_rate = 0.01
n_episodes = 500
start_epsilon = 1.0
epsilon_decay = start_epsilon / (n_episodes / 2)
final_epsilon = 0.0
eps = start_epsilon
discount_factor = 0.95

reward_queue = []

env = gym.make('MountainCar-v0')

Q = defaultdict(lambda: np.zeros(env.action_space.n))

mov_average = []

for episode in range(n_episodes):

    observation = env.reset()
    observation = observation[0]
    observation[0] = round(observation[0],2)
    observation[1] = round(observation[1],2)

    done = False

    print('episode :' + '' + str(episode))

    while not done:
        if np.random.random() < eps:
          action = np.random.randint(0, env.action_space.n)
        else:
          action = np.argmax(Q[(observation[0], observation[1])])

        new_observation, reward, terminated, truncated, info = env.step(action)

        mov_average.append(reward)

        new_observation[0] = round(new_observation[0],2)
        new_observation[1] = round(new_observation[1],2)

        # IF YOU WANT TO CHANGE THE REWARDS
        """
        if new_observation[0] >= 0.5:
          reward = 100
        if new_observation[0] - observation[0] > 0 and action == 2:
            reward = reward + 1
        if new_observation[0] - observation[0] < 0 and action == 0:
            reward = reward + 1
        """

        done = terminated or truncated

        Q[(observation[0],observation[1])][action] += learning_rate * (reward + discount_factor * np.max(Q[(new_observation[0],new_observation[1])]) - Q[(observation[0], observation[1])][action])

        observation = new_observation

    eps = max(final_epsilon, eps - epsilon_decay)
    #print(eps)

print("Exploiting the training...")
reward_queue = moving_average(mov_average, 4020) #201*20

observation = env.reset()
observation = observation[0]
observation[0] = round(observation[0],2)
observation[1] = round(observation[1],2)

done = False

best_observation = observation

while not done:
    action = np.argmax(Q[(observation[0], observation[1])])

    new_observation, reward, terminated, truncated, info = env.step(action)

    new_observation[0] = round(new_observation[0],2)
    new_observation[1] = round(new_observation[1],2)

    if observation[0] >= best_observation[0]:
      best_observation = observation

    done = terminated or truncated

    observation = new_observation

    #env.render()

print(best_observation)

if best_observation[0] >= 0.5:
    print('Flag reached')
else:
   print('Flag not reached')


#PLOT CODE
"""
rolling_length = 500
fig, axs = plt.subplots(ncols=1, figsize=(15, 5))

axs.set_title("Reward Per Episode (Training)")
episode_reward_t = np.convolve(np.array(reward_queue).flatten(), np.ones(rolling_length), mode="valid")/rolling_length
axs.plot(range(len(episode_reward_t)), episode_reward_t)

plt.show()
"""