我正在尝试实现值迭代算法。我有一个网格

grid = [[0, 0, 0, +1],    [0, "W", 0, -1],    [0, 0, 0, 0]]

一个动作列表

actlist = {UP:1, DOWN:2, LEFT:3, RIGHT:4}

和一个奖励函数

reward = [[0, 0, 0, 0],      [0, 0, 0, 0],      [0, 0, 0,  0]] 

我写了一个函数T，它返回一个包含3个元组的元组。

def T(i,j,actions):if(i == 0 and j == 0):    if(actions == UP):        return (i,i,0.8),(i,i,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i+1,j,0.8),(i,j,0.1),(i,j+1,0.1)    elif(actions == LEFT):        return (i,j,0.8),(i,j,0.1),(i+1,j,0.1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i,i,0.1),(i+1,j,0.1)elif (i == 0 and j == 1):    if(actions == UP):        return (i,i,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == LEFT):        return (i,j-1,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i,j,0.1),(i,j,0.1)elif(i == 0 and j == 2):    if(actions == UP):        return (i,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return(i+1,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == LEFT):        return (i,j-1,0.8),(i,j,0.1),(i+1,j,0.1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i,j,0.1),(i+1,j,0.1)elif(i == 0 and j == 3):    if(actions == UP):        return (-1,-1,0.8),(-1,-1,0.1),(-1,-1,0.1)    elif(actions == DOWN):        return (-1,-1,0.8),(-1,-1,0.1),(-1,-1,0.1)    elif(actions == LEFT):        return (-1,-1,0.8),(-1,-1,0.1),(-1,-1,0.1)    elif(actions == RIGHT):        return (-1,-1,0.8),(-1,-1,0.1),(-1,-1,0.1)# 第二行elif (i == 1 and j == 0):    if(actions == UP):        return (i-1,j,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == DOWN):        return (i+1,j,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == LEFT):        return (i,j,0.8),(i-1,j,0.1),(i+1,j,0.1)    elif(actions == RIGHT):        return (i,j,0.8),(i-1,j,0.1),(i+1,j,0.1)elif(i == 1 and j ==1):    if(actions == UP):        return (i,j,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == LEFT):        return (i,j,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == RIGHT):        return (i,j,0.8),(i,j,0.1),(i,j,0.1)elif (i == 1 and j == 2):    if(actions == UP):        return (i-1,j,0.8),(i,j,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i+1,j,0.8),(i,j,0.1),(i,j+1,0.1)    elif(actions == LEFT):        return (i,j,0.8),(i-1,j,0.1),(i+1,j,0.1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i-1,j,0.1),(i+1,j,0.1)elif(i == 1 and j == 3):    if(actions == UP):        return (-2,-2,0.8),(-2,-2,0.1),(-2,-2,0.1)    elif(actions == DOWN):        return (-2,-2,0.8),(-2,-2,0.1),(-2,-2,0.1)    elif(actions == LEFT):        return (-2,-2,0.8),(-2,-2,0.1),(-2,-2,0.1)    elif(actions == RIGHT):        return (-2,-2,0.8),(-2,-2,0.1),(-2,-2,0.1)      # 第三行elif(i == 2 and j == 0):    if(actions == UP):        return (i-1,j,0.8),(i,j,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j,0.1),(i,j+1,1,0.1)    elif(actions == LEFT):        return (i,j,0.8),(i-1,j,0.1),(i,j,0.1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i-1,j,0.1),(i,j,0.1)elif (i == 2 and j == 1):    if(actions == UP):        return (i,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == LEFT):        return (i,j-1,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i,j,0.1),(i,j,0.1)elif(i == 2 and j == 2):    if(actions == UP):        return (i-1,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == LEFT):        return (i,j-1,0.8),(i-1,j,0.1),(i,j,1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i-1,j,0.1),(i,j,0.1)elif(i == 2 and j == 3):    if(actions == UP):        return (i-1,j,0.8),(i,j-1,0.1),(i,j,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j-1,0.1),(i,j,0.1)    elif(actions == LEFT):        return (i,j-1,0.8),(i-1,j,0.1),(i,j,0.1)    elif(actions == RIGHT):        return (i,j,0.8),(i-1,j,0.1),(i,j,0.1) 

这个函数在值迭代函数中被调用：

def value_iteration():U1 = [[0, 0, 0, 0],      [0, 0, 0, 0],      [0, 0, 0, 0]] while True:    U=U1.copy()    delta = 0    for i in range(len(grid)):        for j in range(len(grid[i])):            U1[i][j] = max(sum(p*(R(k,l)+gamma*U[k][l]) for (k,l,p) in T(i,j,a)) for a in actlist)            print(i,j,U1[i][j])            delta = max(delta, abs(U1[i][j] - U[i][j]))    if delta <= epsilon*(1 - gamma)/gamma:        return U

问题是，前两个循环迭代运行正常，输出如下

0 00 10 20 31 01 11 2 1 3 

但之后代码停止并报错

ValueError: too many values to unpack (expected 3)

回答：

查看** **中的元组，可能这就是原因

# 第三行elif(i == 2 and j == 0):    if(actions == UP):        return (i-1,j,0.8),(i,j,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j,0.1),**(i,j+1,1,0.1)**    elif(actions == LEFT):        return (i,j,0.8),(i-1,j,0.1),(i,j,0.1)    elif(actions == RIGHT):        return (i,j+1,0.8),(i-1,j,0.1),(i,j,0.1)elif (i == 2 and j == 1):    if(actions == UP):        return (i,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == DOWN):        return (i,j,0.8),(i,j-1,0.1),(i,j+1,0.1)    elif(actions == LEFT):        return (i,j-1,0.8),(i,j,0.1),(i,j,0.1)    elif(actions == RIGHT):       return (i,j+1,0.8),(i,j,0.1),(i,j,0.1)