def __init__(self):
super().__init__()
self.env = Env.OthelloEnv()
self.initUI()
self.i = 0
self.Q = test.Q
def f(state, action):
e = Env.OthelloEnv()
e.state = state
s, r, d = e._step(action)
return s, d
def main():
value_list = []
env = Env.OthelloEnv()
Value = mc_prediction(random_policy, env, num_episodes=10)
Value[27] = 1
Value[28] = 1
Value[35] = 1
Value[36] = 1
for i in range(64):
value_list.append(Value[i])
# 现有的根据经验得到的矩阵
# value_list = [10, -9, 8, 4, 4, 8, -9, 10,
# -9, -9, -4, -3, -3, -4, -9, -9,
# 8, -4, 8, 2, 2, 8, -4, 8,
# 4, 3, 2, 1, 1, 2, 3, 4,
# 4, 3, 2, 1, 1, 2, 3, 4,
# 8, -4, 8, 2, 2, 8, -4, 8,
# -9, -9, -4, -3, -3, -4, -9, -9,
# 10, -9, 8, 4, 4, 8, -9, 10]
# 打印state矩阵
for i in range(0, 8):
for j in range(0, 8):
print(value_list[8 * i + j], end=' ')
print()
print()
# 下棋开始
def mid_policy(state, flag):
st = judge(state, flag)
l = len(st)
if (l == 0):
return [0, 0]
else:
action = []
for i in range(l):
action.append(value_list[i])
num = max(action)
p = action.index(num)
return [flag, st[p]]
def f(env):
total = 0
s = env._reset()
flag = 1
action = mid_policy(s, flag)
d = False
while not d:
s, r, d = env._step(action)
# for i in range(0,8):
# # for j in range(0,8):
# # if(s[8*i+j]==-1):
# # print(s[8*i+j],end=' ')
# # else:
# # print(s[8*i+j],end=' ')
# # print()
# # print()
flag = -flag
if (flag == 1):
action = mid_policy(s, flag)
else:
action = random_policy(s, flag)
for i in range(0, 8):
for j in range(0, 8):
total += s[8 * i + j]
# print(total)
if (total > 0):
return 1
else:
return 0
env = Env.OthelloEnv()
f(env)
win = 0
for i in range(1000):
win += f(env)
print(win / 1000)
import sys
import random
from collections import defaultdict
import Othello as Env
from Othello import judge
def random_policy(state, flag):
st = judge(state, flag)
l = len(st)
if l == 0:
return [0, 0]
else:
p = random.randint(0, l - 1)
return [flag, st[p]]
def f(state, action):
e = Env.OthelloEnv()
e.state = state
s, r, d = e._step(action)
return s, d
env = Env.OthelloEnv()
env._reset()
ss, dd = f(env.state, random_policy(env.state, 1))
print(ss)
print(dd)