LoginSignup
1
1

More than 5 years have passed since last update.

Codeiqの牛乳配達ルート問題

Last updated at Posted at 2014-11-05

木構造で再帰で解く版

木構造使う練習。

#!/bin/env python
# coding:utf-8

import csv
import sys

"""
根ノードから再帰的に深さ優先で木構造を作成しつつ探索も行う。
コスト集計も中途で行い、最低コストを上回った時点で以降の子孫ノードの探索を中止する。
木構造のプログラムを書くのに慣れていないため、実直にコーディングした。
"""


class Accumulator(object):
    def __init__(self, sum=sys.maxsize, route=[]):
        self.sum = sum
        self.route = route


class Node(object):
    def __init__(self, id, parent=None, cost_from_root=0, children=[]):
        self.id = id
        self.parent = parent
        self.cost_from_root = cost_from_root
        self.children = children

    def __repr__(self):
        return "%i, cost: %i -> %s\n" % (self.id, self.cost_from_root, repr(self.children))


class DeliveryCostCalculator(object):

    def __init__(self, filename):
        self.filename = filename
        self.cost_table = self.get_table()
        self.acc = Accumulator(sys.maxsize, [])

    def get_table(self):
        cost_table = []
        with open(self.filename, 'r') as f:
            reader = csv.reader(f)
            for row in reader:
                cost_table.append([int(col) for col in row])
        return cost_table

    def calc_total_cost(self, current):
        # 残りがなければコスト集計を行う
        tmp = Node(0, current, current.cost_from_root + self.cost_table[current.id][0], None)
        current.children.append(tmp)
        if tmp.cost_from_root < self.acc.sum:
            # コストが最低ならばルートのリストも集計し、アキュムレータに渡す
            self.acc.sum = tmp.cost_from_root

            def _min_r(n, acc):
                if n.parent is None:
                    acc.append(n)
                    return acc
                acc.append(n)
                return _min_r(n.parent, acc)
            self.acc.route = _min_r(tmp, [])
            self.acc.route.reverse()

    def main(self):
        def _f(slot, current):
            if len(slot) <= 0:
                self.calc_total_cost(current)
                return
            for i in slot:
                # 子に次のノードを登録する
                tmp = Node(i, current, current.cost_from_root + self.cost_table[current.id][i])
                if self.acc.sum < tmp.cost_from_root:
                    # この時点でのコストが最低コストを上回っているなら以降の探索を中止する
                    return
                current.children.append(tmp)
                # 追加した数字をリストから削除して再帰する
                a = list(slot)
                a.remove(i)
                _f(a, tmp)
        _f(range(1, len(self.cost_table)), Node(0))
        return self.acc.sum, self.acc.route


if __name__ == "__main__":
    c = DeliveryCostCalculator(sys.argv[1])
    (sum, route) = c.main()
    print(sum)
    print(" -> ".join([str(n.id + 1) for n in route]))
    # print([(n.id + 1, n.cost_from_root) for n in route])

事前に組み合わせを列挙する版

permutationsで繰り返しを許さない順列を作成してあとは普通にforでコスト集計。
集計のところもfunctools.reduceでかっこよくやりたかったけどうまくできなかった。

#!/bin/env python
# coding:utf-8

import csv
import sys
from itertools import permutations

def get_table(filename):
    cost_table = []
    with open(filename, 'r') as f:
        reader = csv.reader(f)
        for row in reader:
            cost_table.append([int(col) for col in row])
    return cost_table


def main(filename):
    cost_table = get_table(filename)
    min_cost = sys.maxsize
    min_route = ()
    for p in permutations(range(1, len(cost_table))):
        # add an initial and a last node
        p = (0,) + p + (0,)
        total_cost = 0
        for i in range(len(p)):
            if i == len(p) - 1:
                continue
            # get a cost between a current and next
            total_cost += cost_table[p[i]][p[i + 1]]
            if min_cost < total_cost:
                break
        # print(total_cost, p)
        if total_cost < min_cost:
            min_cost = total_cost
            min_route = p
    return min_cost, min_route


if __name__ == "__main__":
    c, r = main(sys.argv[1])
    print(c)
    print(" -> ".join([str(n + 1) for n in r]))
1
1
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
1
1