Problem UVA11134-Fabled Rooks

Accept: 716  Submit: 6134
Time Limit: 3000mSec

Problem Description

We would like to place n rooks, 1 ≤ n ≤ 5000, on a n × n board subject to the following restrictions
• The i-th rook can only be placed within the rectangle given by its left-upper corner (xli,yli) and its rightlower corner (xri,yri), where 1 ≤ i ≤ n, 1 ≤ xli ≤ xri ≤ n, 1 ≤ yli ≤ yri ≤ n.

• No two rooks can attack each other, that is no two rooks can occupy the same column or the same row.

Input

The input consists of several test cases. The first line of each of them contains one integer number, n, the side of the board. n lines follow giving the rectangles where the rooks can be placed as described above. The i-th line among them gives xli, yli, xri, and yri. The input file is terminated with the integer ‘0’ on a line by itself.

 Output

Your task is to find such a placing of rooks that the above conditions are satisfied and then output n lines each giving the position of a rook in order in which their rectangles appeared in the input. If there are multiple solutions, any one will do. Output ‘IMPOSSIBLE’ if there is no such placing of the rooks.

 Sample Input

1 1 2 2
5 7 8 8
2 2 5 5
2 2 5 5
6 3 8 6
6 3 8 5
6 3 8 8
3 6 7 8
8
1 1 2 2
5 7 8 8
2 2 5 5
2 2 5 5
6 3 8 6
6 3 8 5
6 3 8 8
3 6 7 8
0
 
题解:有了lrj前面几道关于区间的例题,这个题的贪心还是很好想的,首先因为行列独立,因此可以转化为两个问题分析,在区间1~n之间选n个互不相同的数,使得第i个数在区间[ai,bi].区间排序时按右端点递增排,右端点相同时按左端点从大到小排。每次选取区间最左边能选的点。当有一个区间不能选时就是IMPOSSIBLE.我粗略地证明一下:假设在当前区间不选可选的最左边的点,必定选取了一个更靠右的点,最左边的点肯定也是要填的,之后填这个格子的区间如果是左端点小于等于当前区间的区间,那么交换填的方式不影响可解性,如果左端点大于当前区间左端点,那么当前区间尽量向左填所得到的解不会比当前解差。有不对的地方,欢迎大家指教。
 
 #include <bits/stdc++.h>

 using namespace std;

 const int maxn =  + ;

 int n;

 struct Interval {
pair<int, int> interval;
int pos;
}hor[maxn], ver[maxn]; bool cmp(const Interval &a, const Interval &b) {
if (a.interval.second == b.interval.second) return a.interval.first > b.interval.first;
else return a.interval.second < b.interval.second;
} bool hvis[maxn], vvis[maxn];
int vans[maxn], hans[maxn]; int main()
{
//freopen("input.txt", "r", stdin);
while (~scanf("%d", &n) && n) {
memset(hvis, false, sizeof(hvis));
memset(vvis, false, sizeof(vvis));
int xl, yl, xr, yr;
for (int i = ; i <= n; i++) {
scanf("%d%d%d%d", &xl, &yl, &xr, &yr);
ver[i].interval = make_pair(xl, xr);
ver[i].pos = i;
hor[i].interval = make_pair(yl, yr);
hor[i].pos = i;
} sort(ver + , ver + + n, cmp);
sort(hor + , hor + + n, cmp); bool ok = true; for (int i = ; i <= n; i++) {
int l = ver[i].interval.first, r = ver[i].interval.second;
int p = ver[i].pos;
int j;
for (j = l; j <= r; j++) {
if (!vvis[j]) {
vvis[j] = true;
vans[p] = j;
break;
}
}
if (j == r + ) { ok = false; break; }
} if (!ok) {
printf("IMPOSSIBLE\n");
continue;
} for (int i = ; i <= n; i++) {
int l = hor[i].interval.first, r = hor[i].interval.second;
int p = hor[i].pos;
int j;
for (j = l; j <= r; j++) {
if (!hvis[j]) {
hvis[j] = true;
hans[p] = j;
break;
}
}
if (j == r + ) { ok = false; break; }
} if (!ok) {
printf("IMPOSSIBLE\n");
continue;
} for (int i = ; i <= n; i++) {
printf("%d %d\n", vans[i], hans[i]);
}
}
return ;
}
 

 Sample Output

Problem UVA11134-Fabled Rooks(贪心)的更多相关文章

  1. UVA-11134 Fabled Rooks 贪心问题(区间贪心)

    题目链接:https://cn.vjudge.net/problem/UVA-11134 题意 在 n*n 的棋盘上,放上 n 个车(ju).使得这 n 个车互相不攻击,即任意两个车不在同一行.同一列 ...

  2. uva11134 - Fabled Rooks(问题分解,贪心法)

    这道题非常好,不仅用到了把复杂问题分解为若干个熟悉的简单问题的方法,更是考察了对贪心法的理解和运用是否到位. 首先,如果直接在二维的棋盘上考虑怎么放不好弄,那么注意到x和y无关(因为两个车完全可以在同 ...

  3. UVA - 11134 Fabled Rooks[贪心 问题分解]

    UVA - 11134 Fabled Rooks We would like to place n rooks, 1 ≤ n ≤ 5000, on a n × n board subject to t ...

  4. UVA 11134 - Fabled Rooks(贪心+优先队列)

    We would like to place  n  rooks, 1 ≤  n  ≤ 5000, on a  n×n  board subject to the following restrict ...

  5. Uva11134 Fabled Rooks

    普通的贪心题. 虽然图是二维的,但可以把横向和纵向分开处理. 将区间按右端点排序,然后从区间左端点到右端点找第一个空位置放棋子即可. /*by SilverN*/ #include<algori ...

  6. UVA 11134 Fabled Rooks 贪心

    题目链接:UVA - 11134 题意描述:在一个n*n(1<=n<=5000)的棋盘上放置n个车,每个车都只能在给定的一个矩形里放置,使其n个车两两不在同一行和同一列,判断并给出解决方案 ...

  7. L - Fabled Rooks(中途相遇法和贪心)

    Problem F: Fabled Rooks We would like to place n rooks, 1 ≤ n ≤ 5000, on a n×n board subject to the ...

  8. 贪心 uvaoj 11134 Fabled Rooks

    Problem F: Fabled Rooks We would like to place n rooks, 1 ≤ n ≤ 5000, on a n×n board subject to the ...

  9. 01_传说中的车(Fabled Rooks UVa 11134 贪心问题)

    问题来源:刘汝佳<算法竞赛入门经典--训练指南> P81: 问题描述:你的任务是在n*n(1<=n<=5000)的棋盘上放n辆车,使得任意两辆车不相互攻击,且第i辆车在一个给定 ...

  10. UVA 11134 Fabled Rooks(贪心的妙用+memset误用警示)

    题目链接: https://cn.vjudge.net/problem/UVA-11134 /* 问题 输入棋盘的规模和车的数量n(1=<n<=5000),接着输入n辆车的所能在的矩阵的范 ...

随机推荐

  1. 【Java每日一题】20170302

    20170301问题解析请点击今日问题下方的“[Java每日一题]20170302”查看(问题解析在公众号首发,公众号ID:weknow619) package Mar2017; public cla ...

  2. Java高并发--AQS

    Java高并发--AQS 主要是学习慕课网实战视频<Java并发编程入门与高并发面试>的笔记 AQS是AbstractQueuedSynchronizer的简称,直译过来是抽象队列同步器. ...

  3. PHP 无限极分类下拉列表实现

    1. 递归实现下拉列表 /** * @param 递归 实现下拉列表 分类 */ include('db.inc.php'); function getList($pid = 0,&$resu ...

  4. .net core 下编码问题

    System.Globalization.CultureInfo.CurrentCulture = new System.Globalization.CultureInfo("zh-CN&q ...

  5. Python函数式编程(二):常见高级函数

    一个函数的参数中有函数作为参数,这个函数就为高级函数. 下面学习几个常见高级函数. ---------------------------------------------------------- ...

  6. Android gravity和layout_gravity的区别

    一.gravity和layout_gravity相同处 两者都是设置对齐方式的属性.内部的属性值相同. 根据英文意思也能理解其中的意思.如center_horizontal表示在水平方向上的位置为中间 ...

  7. 使用Visual Studio Team Services持续集成(一)——构建ASP.NET Core

    使用Visual Studio Team Services持续集成(一)--构建ASP.NET Core 概述 持续集成(CI)是将代码尽可能频繁地集成到共享仓库中的过程.在代码集成期间,构建中断或测 ...

  8. NAT穿越(一) NAT类型

    NAT分为四种类型: (1)完全透明NAT(Full Cone NAT): 从内部主机  (IN IP ipa) +端口(IN PORT porta) 发送的数据映射为  IP(OUT IP IPA) ...

  9. MySQL优化技巧【持续更新】

    前言 应用程序或web网页有时慢的像蜗牛爬似的,可能是网络原因,可能是系统架构原因,还有可能是数据库原因.那么如何提高数据库SQL语句执行速度呢?下面是积累的一些优化技巧,望对君有用. 正文 1.比较 ...

  10. ORACLE实际执行计划与预估执行计划不一致性能优化案例

      在一台ORACLE服务器上做巡检时,使用下面SQL找出DISK_READ最高的TOP SQL分析时,分析过程中,有一条SQL语句的一些反常现象,让人觉得很奇怪: SELECT SQL_ID,    ...