Python 生成器在 minimax 的递归实现中出现问题

Question

我正在研究 minimax 算法的实现来解决 2048。我在遍历对象列表并将此值传递给递归函数时遇到问题。代码是：

def MAXIMIZE(state,deep):
if deep == 10:
    return (None,greed(state))

tup = (None, -1000000) #maxChild, maxUtility 
moves = state.getAvailableMoves()
#children by moves
children = []
children.append(state.clone().move(m) for m in moves)
for child in children:
    temp,utility = MINIMIZE(child,deep + 1)
    if utility > tup[1]:
        tup = (child, utility)

return tup
def MINIMIZE(state,deep):
if deep == 10:
    return (None, greed(state))
tup = (None, +10000000)

cells = state.getAvailableCells() # this is where I get an error - "'generator' object has no attribute 'getAvailableCells'"

children = []
children.append(state.clone().setCellValue(cell, 2) for cell in cells)
for child in children:
    temp,utility = MAXIMIZE(child,deep + 1)
    if utility < tup[1]:
        tup = (child,utility)
return tup

在 MINIMIZE 函数中，行 - cells = state.getAvailableCells() 给出“'generator' object has no attribute 'getAvailableCells'” 谁能帮我解决这个问题？（我是学物理的，对Python的了解有限，翻了各种现成的题，也没看懂。）

Answer 1

这里

children = []
children.append(state.clone().move(m) for m in moves)

您创建一个空 list，然后向其中添加一个元素，该元素是一个生成器（由 state.clone().move(m) for m in moves generator expression 创建）。

您想要的是列表表达式 - 将这两行替换为：

children = [state.clone().move(m) for m in moves)]

这将评估生成器并从中构建和填充列表。

请注意，您在这里遇到了同样的问题：

children = []
children.append(state.clone().setCellValue(cell, 2) for cell in cells)

需要相同的修复：

children = [state.clone().setCellValue(cell, 2) for cell in cells]

Python 生成器在 minimax 的递归实现中出现问题

Trouble with Python generators in recursive implementation of minimax

python

recursion

generator

minimax