def update():
for episode in range(100):
# initial observation
observation = env.reset()
while True:
env.render()
# RL choose action based on observation
action = RL.choose_action(str(observation))
# RL take action and get next observation and reward
observation_, reward, done = env.step(action)
# RL learn from this transition
RL.learn(str(observation), action, reward, str(observation_))
# swap observation
observation = observation_
if done:
break
# end of navigation
print('Reinforcement Learning done successful')
env.destroy()
if __name__ == "__main__":
env = Maze()
RL = QLearningTable(actions=list(range(env.n_actions)))
env.after(100, update)
env.mainloop()
step += 1
# end of game
print('game over')
env.destroy()
if __name__ == "__main__":
# maze game
env = Maze()
# RL = DeepQNetwork(env.n_actions, env.n_features,
# learning_rate=0.01,
# reward_decay=0.9,
# e_greedy=0.9,
# replace_target_iter=200,
# memory_size=2000
# )
# param tuning by hand, best version for now
RL = DeepQNetwork(env.n_actions,
env.n_features,
learning_rate=0.005,
reward_decay=0.8,
e_greedy=0.8,
replace_target_iter=200,
memory_size=2000)
env.after(100, run_maze)
env.mainloop()
RL.plot_cost()
exit()
# =============================================================================
# if (episode+1) >= 3:
# action = RL.choose_ma_action(str(observation))
# else:
# action = RL.choose_action(str(observation))
# =============================================================================
#action = RL.choose_ma_action(str(observation))
action = RL.choose_entropy_action(str(observation))
#action = RL.choose_action(str(observation))
observation_, reward, done = env.step(action)
total_reward_value = total_reward_value + reward
RL.learn(str(observation), action, reward, str(observation_))
observation = observation_
if done:
break
print('{} episode over'.format(episode + 1))
print('average reward {}'.format(total_reward_value / (episode + 1)))
average_reward.append(total_reward_value / (episode + 1))
#print(RL.q_table)
#RL.q_table.to_csv('3-2.csv',header=True,index=True)
print(average_reward)
env.destroy()
#RL.q_table.to_clipboard()
if __name__ == '__main__':
env = Maze()
RL = QLearningTable(actions=list(range(env.n_actions)))
env.after(2000, update)
env.mainloop()