强化学习#code3

把后面的版本v0改成v1，random_policy=np.ones((env.observation_space.n,env.action_space.n))/env.action_space.n
就能运行出来，新版本有bug. >_<

import numpy as np
import gym
np.random.seed(0)
env = gym.make('FrozenLake-v1')
env = env.unwrapped  # 据说不做这个动作会有很多限制，unwrapped是打开限制的意思
print(env.action_space)#env.action_space查看这个环境中可用的action有多少个，返回Discrete()格式
print(env.observation_space)#env.observation_space查看这个环境中可用的observation有多少个，返回Discrete()格式
np.random.seed(0)#使得随机数据可预测
env.seed(0)# 设置随机数种子，只是为了让结果可以精确复现，一般情况下可删去
env.unwrapped.P[14][2] # 查看动力
print('观察空间 = {}'.format(env.observation_space))
print('动作空间 = {}'.format(env.action_space))
print('观测空间大小 = {}'.format(env.observation_space.n))
print('动作空间大小 = {}'.format(env.action_space.n))
#用随机策略玩玩
def play_policy(env, policy, render=False):
    total_reward = 0.
    observation = env.reset()
    while True:
        if render:
            env.render() # 此行可显示
        action = np.random.choice(env.action_space.n,
                p=policy[observation])
        observation, reward, done, _ = env.step(action)
        total_reward += reward  # 统计回合奖励
        if done: # 游戏结束
            break
    return total_reward
# 随机策略
random_policy = \
        np.ones((env.observation_space.n,env.action_space.n)) / env.action_space.n
        #np.ones((env.unwrapped.nS, env.unwrapped.nA)) / env.unwrapped.nA   
episode_rewards = [play_policy(env, random_policy)  for _ in range(100)]
print("随机策略 平均奖励：{}".format(np.mean(episode_rewards)))
#--------------策略评估-------------
def v2q(env, v, s=None, gamma=1.): # 根据状态价值函数计算动作价值函数
    if s is not None: # 针对单个状态求解
        q = np.zeros(env.action_space.n)
        for a in range(env.action_space.n):
            for prob, next_state, reward, done in env.unwrapped.P[s][a]:
                q[a] += prob * \
                        (reward + gamma * v[next_state] * (1. - done))
    else: # 针对所有状态求解
        q = np.zeros((env.observation_space.n,env.action_space.n))
        for s in range(env.observation_space.n):
            q[s] = v2q(env, v, s, gamma)
    return q
 
def evaluate_policy(env, policy, gamma=1., tolerant=1e-6):
    v = np.zeros(env.observation_space.n) # 初始化状态价值函数
    while True: # 循环
        delta = 0
        for s in range(env.observation_space.n):
            vs = sum(policy[s] * v2q(env, v, s, gamma)) # 更新状态价值函数
            delta = max(delta, abs(v[s]-vs)) # 更新最大误差
            v[s] = vs # 更新状态价值函数
        if delta < tolerant: # 查看是否满足迭代条件
            break
    return v
#对随机策略进行策略进行策略评估
print('状态价值函数：')
v_random = evaluate_policy(env, random_policy)
print(v_random.reshape(4, 4))
 
print('动作价值函数：')
q_random = v2q(env, v_random)
print(q_random)
#---------------策略改进-----------------------
def improve_policy(env, v, policy, gamma=1.):
    optimal = True
    for s in range(env.observation_space.n):
        q = v2q(env, v, s, gamma)
        a = np.argmax(q)
        if policy[s][a] != 1.:
            optimal = False
            policy[s] = 0.
            policy[s][a] = 1.
    return optimal
#对随机策略进行改进
policy = random_policy.copy()
optimal = improve_policy(env, v_random, policy)
if optimal:
    print('无更新，最优策略为：')
else:
    print('有更新，更新后的策略为：')
print(policy)
#-------------策略迭代--------------
def iterate_policy(env, gamma=1., tolerant=1e-6):
     # 初始化为任意一个策略
    policy = np.ones((env.observation_space.n,env.action_space.n)) \
            / env.action_space.n
    while True:
        v = evaluate_policy(env, policy, gamma, tolerant) # 策略评估
        if improve_policy(env, v, policy): # 策略改进
            break
    return policy, v
 
policy_pi, v_pi = iterate_policy(env)
print('状态价值函数 =')
print(v_pi.reshape(4, 4))
print('最优策略 =')
print(np.argmax(policy_pi, axis=1).reshape(4, 4))
def iterate_policy(env, gamma=1., tolerant=1e-6):
     # 初始化为任意一个策略
    policy = np.ones((env.observation_space.n,env.action_space.n)) \
            / env.action_space.n
    while True:
        v = evaluate_policy(env, policy, gamma, tolerant) # 策略评估
        if improve_policy(env, v, policy): # 策略改进
            break
    return policy, v
 
policy_pi, v_pi = iterate_policy(env)
print('状态价值函数 =')
print(v_pi.reshape(4, 4))
print('最优策略 =')
print(np.argmax(policy_pi, axis=1).reshape(4, 4))
#------------测试策略----------
episode_rewards = [play_policy(env, policy_pi)  for _ in range(100)]
print("策略迭代 平均奖励：{}".format(np.mean(episode_rewards)))
#------------价值迭代------------
def iterate_value(env, gamma=1, tolerant=1e-6):
    v = np.zeros(env.observation_space.n) # 初始化
    while True:
        delta = 0
        for s in range(env.observation_space.n):
            vmax = max(v2q(env, v, s, gamma)) # 更新价值函数
            delta = max(delta, abs(v[s]-vmax))
            v[s] = vmax
        if delta < tolerant: # 满足迭代需求
            break
            
    policy = np.zeros((env.observation_space.n,env.action_space.n)) # 计算最优策略
    for s in range(env.observation_space.n):
        a = np.argmax(v2q(env, v, s, gamma))
        policy[s][a] = 1.
    return policy, v
 
policy_vi, v_vi = iterate_value(env)
print('状态价值函数 =')
print(v_vi.reshape(4, 4))
print('最优策略 =')
print(np.argmax(policy_vi, axis=1).reshape(4, 4))
#-------------测试随机策略---------
episode_rewards = [play_policy(env, policy_vi) for _ in range(100)]
print("价值迭代 平均奖励：{}".format(np.mean(episode_rewards)))

结果

Discrete(4)
Discrete(16)
观察空间 = Discrete(16)
动作空间 = Discrete(4)
观测空间大小 = 16
动作空间大小 = 4
随机策略 平均奖励：0.0
状态价值函数：
[[0.0139372  0.01162942 0.02095187 0.01047569]
 [0.01624741 0.         0.04075119 0.        ]
 [0.03480561 0.08816967 0.14205297 0.        ]
 [0.         0.17582021 0.43929104 0.        ]]
动作价值函数：
[[0.01470727 0.01393801 0.01393801 0.01316794]
 [0.00852221 0.01162969 0.01086043 0.01550616]
 [0.02444416 0.0209521  0.02405958 0.01435233]
 [0.01047585 0.01047585 0.00698379 0.01396775]
 [0.02166341 0.01701767 0.0162476  0.01006154]
 [0.         0.         0.         0.        ]
 [0.05433495 0.04735099 0.05433495 0.00698396]
 [0.         0.         0.         0.        ]
 [0.01701767 0.04099176 0.03480569 0.04640756]
 [0.0702086  0.11755959 0.10595772 0.05895286]
 [0.18940397 0.17582024 0.16001408 0.04297362]
 [0.         0.         0.         0.        ]
 [0.         0.         0.         0.        ]
 [0.08799662 0.20503708 0.23442697 0.17582024]
 [0.25238807 0.53837042 0.52711467 0.43929106]
 [0.         0.         0.         0.        ]]
有更新，更新后的策略为：
[[1. 0. 0. 0.]
 [0. 0. 0. 1.]
 [1. 0. 0. 0.]
 [0. 0. 0. 1.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [0. 0. 0. 1.]
 [0. 1. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [0. 0. 1. 0.]
 [0. 1. 0. 0.]
 [1. 0. 0. 0.]]
状态价值函数 =
[[0.82351246 0.82350689 0.82350303 0.82350106]
 [0.82351416 0.         0.5294002  0.        ]
 [0.82351683 0.82352026 0.76469786 0.        ]
 [0.         0.88234658 0.94117323 0.        ]]
最优策略 =
[[0 3 3 3]
 [0 0 0 0]
 [3 1 0 0]
 [0 2 1 0]]
状态价值函数 =
[[0.82351246 0.82350689 0.82350303 0.82350106]
 [0.82351416 0.         0.5294002  0.        ]
 [0.82351683 0.82352026 0.76469786 0.        ]
 [0.         0.88234658 0.94117323 0.        ]]
最优策略 =
[[0 3 3 3]
 [0 0 0 0]
 [3 1 0 0]
 [0 2 1 0]]
策略迭代 平均奖励：0.73
状态价值函数 =
[[0.82351232 0.82350671 0.82350281 0.82350083]
 [0.82351404 0.         0.52940011 0.        ]
 [0.82351673 0.82352018 0.76469779 0.        ]
 [0.         0.88234653 0.94117321 0.        ]]
最优策略 =
[[0 3 3 3]
 [0 0 0 0]
 [3 1 0 0]
 [0 2 1 0]]
PS D:\python\code\book\QL>  d:; cd 'd:\python\code\book\QL'; & 'D:\python\python.exe' 'c:\Users\申勤勤\.vscode\extensions\ms-python.python-2022.2.1924087327\pythonFiles\lib\python\debugpy\launcher' '61607' '--' 'd:\python\code\book\QL\Frozen.py' 
Discrete(4)
Discrete(16)
观察空间 = Discrete(16)
动作空间 = Discrete(4)
观测空间大小 = 16
动作空间大小 = 4
随机策略 平均奖励：0.0
状态价值函数：
[[0.0139372  0.01162942 0.02095187 0.01047569]
 [0.01624741 0.         0.04075119 0.        ]
 [0.03480561 0.08816967 0.14205297 0.        ]
 [0.         0.17582021 0.43929104 0.        ]]
动作价值函数：
[[0.01470727 0.01393801 0.01393801 0.01316794]
 [0.00852221 0.01162969 0.01086043 0.01550616]
 [0.02444416 0.0209521  0.02405958 0.01435233]
 [0.01047585 0.01047585 0.00698379 0.01396775]
 [0.02166341 0.01701767 0.0162476  0.01006154]
 [0.         0.         0.         0.        ]
 [0.05433495 0.04735099 0.05433495 0.00698396]
 [0.         0.         0.         0.        ]
 [0.01701767 0.04099176 0.03480569 0.04640756]
 [0.0702086  0.11755959 0.10595772 0.05895286]
 [0.18940397 0.17582024 0.16001408 0.04297362]
 [0.         0.         0.         0.        ]
 [0.         0.         0.         0.        ]
 [0.08799662 0.20503708 0.23442697 0.17582024]
 [0.25238807 0.53837042 0.52711467 0.43929106]
 [0.         0.         0.         0.        ]]
有更新，更新后的策略为：
[[1. 0. 0. 0.]
 [0. 0. 0. 1.]
 [1. 0. 0. 0.]
 [0. 0. 0. 1.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [0. 0. 0. 1.]
 [0. 1. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [1. 0. 0. 0.]
 [0. 0. 1. 0.]
 [0. 1. 0. 0.]
 [1. 0. 0. 0.]]
状态价值函数 =
[[0.82351246 0.82350689 0.82350303 0.82350106]
 [0.82351416 0.         0.5294002  0.        ]
 [0.82351683 0.82352026 0.76469786 0.        ]
 [0.         0.88234658 0.94117323 0.        ]]
最优策略 =
[[0 3 3 3]
 [0 0 0 0]
 [3 1 0 0]
 [0 2 1 0]]
状态价值函数 =
[[0.82351246 0.82350689 0.82350303 0.82350106]
 [0.82351416 0.         0.5294002  0.        ]
 [0.82351683 0.82352026 0.76469786 0.        ]
 [0.         0.88234658 0.94117323 0.        ]]
最优策略 =
[[0 3 3 3]
 [0 0 0 0]
 [3 1 0 0]
 [0 2 1 0]]
策略迭代 平均奖励：0.73
状态价值函数 =
[[0.82351232 0.82350671 0.82350281 0.82350083]
 [0.82351404 0.         0.52940011 0.        ]
 [0.82351673 0.82352018 0.76469779 0.        ]
 [0.         0.88234653 0.94117321 0.        ]]
最优策略 =
[[0 3 3 3]
 [0 0 0 0]
 [3 1 0 0]
 [0 2 1 0]]
价值迭代 平均奖励：0.84

现在懒得整理了，第三组讲完再做叭，先把第四章的图跑出来！！！