heapq 自定义 compareTo

heapq custom compareTo

我正在尝试定义一个自定义方法来将有序插入到 python 中的优先级队列中,但没有得到预期的结果。一旦将插入方法定义到队列中,如下所示:

def insert(self, node):
    if isinstance(node, treeNode):
        heapq.heappush(self._frontier, (node._f, node))
    else:
        print("Error. It is not a node.")

并在 'node' class 中实施以下 lt

def __lt__(self, other):
    return self._f < other._f

插入不是由 'f' 属性值完成的,这是我想做的,按该值确定的升序插入。任何帮助将不胜感激。

失败示例:

[(141.09530289033592, <treeNode.treeNode object at 0x7f08bb840fd0>), (484.8315227978057, <treeNode.treeNode object at 0x7f08bb849da0>), (390.0514031446352, <treeNode.treeNode object at 0x7f08bb840e48>)]

它只将最低值放在第一个位置,这确实有意义,因为使用了优先级队列,但后面的不是按我想声明的自定义方法排序的。

正如@Bakuriu 提到的,heapq 仅保持 heap, if you want to obtain the elements in order use nsmallest 的不变量,例如:

import heapq


class TreeNode:
    def __init__(self, f):
        self.f = f

    def __repr__(self):
        return 'f: {}'.format(self.f)

    def __lt__(self, other):
        return self.f < other.f


class Frontier:
    def __init__(self):
        self.frontier = []

    def insert(self, node):
        if isinstance(node, TreeNode):
            heapq.heappush(self.frontier, (node.f, node))
        else:
            print("Error. It is not a node.")

    def __len__(self):
        return len(self.frontier)


t = Frontier()

for n in [TreeNode(141.09530289033592), TreeNode(484.8315227978057), TreeNode(390.0514031446352)]:
    t.insert(n)

print(heapq.nsmallest(len(t), t.frontier))

输出

[(141.09530289033592, f: 141.09530289033592), (390.0514031446352, f: 390.0514031446352), (484.8315227978057, f: 484.8315227978057)]