Question

我正在尝试使用迭代深化深度优先搜索来创建RushHour拼图解算器。现在，当我执行脚本时，它会生成太多节点并快速占用整个内存（分配速度大约为20MB / s）。我知道我存储板状态的解决方案并不完美但不应该吃太多内存。（整个项目可以在这里找到 here）

import map from node import Node


def all_moves(node):  # returns all valid moves for node
    state = node.state
    output = []
    for i in state.cars:
        car_id = i[0]
        y = i[1]
        x = i[2]
        for action in actions:
            movement = state.is_movable(action, y, x)
            if movement[1] != 0:
                movement.append(car_id)
                movement.append(action)
                output.append(movement)
    return output


def generate_all_child(node):  # creates every new state based on list of valid moves and makes it a child
    for move in all_moves(node):
        action = move[-1]
        car = move[0]
        car_id = move[-2]
        for step in range(1, move[1]+1):
            child_state = node.state.action(car_id, car, action, step)
            child_node = Node(child_state, node)
            node.add_child(child_node)


def dls(node, node_depth):
    if node_depth >= 0:
        generate_all_child(node)
        if node.state.is_solved():
            return node
        for child in node.children:
            dls(child, node_depth-1)


stack = [] actions = ["left", "right", "up", "down"]

mapa = map.Map() mapa.load("krizovatka.csv") mapa.display_matrix() mapa.load_cars() root = Node(mapa, None) stack.append(root)

depth = 5 for i in range(1, depth+1):
    print(i)
    final = dls(root, i)
    print(len(root.children))

Answer 1

问题是is_solved（）（map.py）中的错误。即使最终状态发生，它也返回0，所以它永远不会停止搜索。现在它有效，虽然速度很慢，但这是一个不同的问题。

Python3 RushHour求解器使用迭代加深

1 个答案: