A well-known art union called "Kalevich is Alive!" manufactures objects d'art (pictures). The union consists of n painters who decided to organize their work as follows.

Each painter uses only the color that was assigned to him. The colors are distinct for all painters. Let's assume that the first painter uses color 1, the second one uses color 2, and so on. Each picture will contain all these n colors. Adding the j-th color to the i-th picture takes the j-th painter tij units of time.

Order is important everywhere, so the painters' work is ordered by the following rules:

  • Each picture is first painted by the first painter, then by the second one, and so on. That is, after the j-th painter finishes working on the picture, it must go to the (j + 1)-th painter (if j < n);
  • each painter works on the pictures in some order: first, he paints the first picture, then he paints the second picture and so on;
  • each painter can simultaneously work on at most one picture. However, the painters don't need any time to have a rest;
  • as soon as the j-th painter finishes his part of working on the picture, the picture immediately becomes available to the next painter.

Given that the painters start working at time 0, find for each picture the time when it is ready for sale.

Input

The first line of the input contains integers m, n (1 ≤ m ≤ 50000, 1 ≤ n ≤ 5), where m is the number of pictures and n is the number of painters. Then follow the descriptions of the pictures, one per line. Each line contains n integers ti1, ti2, ..., tin (1 ≤ tij ≤ 1000), where tij is the time the j-th painter needs to work on the i-th picture.

Output

Print the sequence of m integers r1, r2, ..., rm, where ri is the moment when the n-th painter stopped working on the i-th picture.

Examples
input

Copy
5 1
1
2
3
4
5
output

Copy
1 3 6 10 15 
input

Copy
4 2
2 5
3 1
5 3
10 1
output

Copy
7 8 13 21 

做题像是憋屎,想明白后,一下就出来了。
#include <iostream>
#include <vector>
#include <algorithm>
#include <string>
#include <set>
#include <queue>
#include <map>
#include <sstream>
#include <cstdio>
#include <cstring>
#include <numeric>
#include <cmath>
#include <unordered_set>
#include <unordered_map>
#include <xfunctional>
#define ll long long
#define mod 998244353
using namespace std;
int dir[][] = { {,},{,-},{-,},{,} };
const int maxn = ;
const long long inf = 0x7f7f7f7f7f7f7f7f; int main()
{
int m, n;
cin >> m >> n;
vector<vector<int>> t(m+, vector<int>(n+,));
for (int i = ; i <= m; i++)
for (int j = ; j <= n; j++)
cin >> t[i][j];
for (int i = ; i <= m; i++)
{
for (int j = ; j <= n; j++)
{
if (t[i][j - ] < t[i - ][j])
t[i][j] += t[i - ][j];
else
t[i][j] += t[i][j - ];
}
}
for (int i = ; i <= m; i++)
{
cout << t[i][n] << " ";
}
return ;
}

Art Union的更多相关文章

  1. Codeforces Round #241 (Div. 2)->B. Art Union

    B. Art Union time limit per test 1 second memory limit per test 256 megabytes input standard input o ...

  2. Codeforces Round #241 (Div. 2) B. Art Union 基础dp

    B. Art Union time limit per test 1 second memory limit per test 256 megabytes input standard input o ...

  3. Codeforces Round #241 (Div. 2) B. Art Union (DP)

    题意:有\(n\)个画家,\(m\)幅画,每个画家负责\(m\)幅画,只有前一个画家画完时,后面一个画家才能接着画,一个画家画完某幅画的任务后,可以开始画下一幅画的任务,问每幅画最后一个任务完成时的时 ...

  4. n个人作m幅画

    题目网址: http://codeforces.com/problemset/problem/416/B A well-known art union called "Kalevich is ...

  5. Codeforces Round #241 (Div. 2) B dp

    B. Art Union time limit per test 1 second memory limit per test 256 megabytes input standard input o ...

  6. The Art of Deception

    前言 一些黑客毁坏别人的文件甚至整个硬盘,他们被称为电脑狂人(crackers)或计算机破坏者(vandals).另一些新手省去学习技术的麻烦,直接下载黑客工具侵入别人的计算机,这些人被称为脚本小子( ...

  7. Android内存优化(一)DVM和ART原理初探

    相关文章 Android内存优化系列 Java虚拟机系列 前言 要学习Android的内存优化,首先要了解Java虚拟机,此前我用了多篇文章来介绍Java虚拟机的知识,就是为了这个系列做铺垫.在And ...

  8. SQL Server-聚焦UNIOL ALL/UNION查询(二十三)

    前言 本节我们来看看有关查询中UNION和UNION ALL的问题,简短的内容,深入的理解,Always to review the basics. 初探UNION和UNION ALL 首先我们过一遍 ...

  9. SQL 提示介绍 hash/merge/concat union

    查询提示一直是个很有争议的东西,因为他影响了sql server 自己选择执行计划.很多人在问是否应该使用查询提示的时候一般会被告知慎用或不要使用...但是个人认为善用提示在不修改语句的条件下,是常用 ...

随机推荐

  1. VAE

    Waiting list: basic knowledge: http://adamlineberry.ai/vae-series/vae-code-experiments

  2. vue(六)--计算属性(computed)

    计算属性关键词: computed demo1: <div id="app"> <p>原始字符串: {{ message }}</p> < ...

  3. Spring AOP-基于@AspectJ风格

    关于Spring AOP,可以去看看官方文档: https://docs.spring.io/spring-framework/docs/current/spring-framework-refere ...

  4. 完成:vue-styled-components

    基础 https://es6.ruanyifeng.com/#docs/string#模板字符串 注释:模板字符串返回字符串类型 如果${...}中的值不是字符串,将按照一般的规则转为字符串.比如,大 ...

  5. C# 图片转为Base64

    /// <summary> /// 图片转Base64 /// </summary> /// <param name="ImageFileName"& ...

  6. liner-classifiers-SVM

    1支持向量机 参考看了这篇文章你还不懂SVM你就来打我 第一遍看完确实有想打死作者的冲动,但是多看几遍之后,真香~ [SVM---这可能是最直白的推导了] 个人觉得这篇文章讲的很清楚,条理清晰,数学推 ...

  7. 神经网络反向传播算法&&卷积神经网络

    听一遍课程之后,我并不太明白这个算法的奇妙之处?? 为啥? 神经网络反向传播算法 神经网络的训练依靠反向传播算法,最开始输入层输入特征向量,网络层计算获得输出,输出层发现输出和正确的类号不一样,这时就 ...

  8. python 的eval函数

    python中的eval()函数是用来计算所有数学的代数计算式,这样可以很快得到复杂代数式的结果. 例如:383660347*375705824-1796136991-1726898699*18994 ...

  9. HDU5608

    题意 英文 做法 设\(g(n)=n^2-3n+2\),有\(g(n)=\sum\limits_{d|n}f(d)\),反演一下有\(f(n)=\sum\limits_{d|n}\mu(\frac{n ...

  10. 野路子码农系列(8)我终于大致搞懂了GBDT

    由于下下周要在组里介绍一个算法,最近开始提前准备,当初非常自信地写下自己最喜欢的GBDT,但随着逐步深入,发现其实自己对这个算法的细节并不是非常了解,了解的只是一些面试题的答案而已……(既然没有深入了 ...