【暑假】[深入动态规划]UVa 1380 A Scheduling Problem
UVa 1380 A Scheduling Problem
题目:
http://acm.hust.edu.cn/vjudge/problem/viewProblem.action?id=41557
思路:
给出一个任务调度树,单向边u->v表示u必须在v之前完成,双向边u-v表示无所谓方向。
题目给出定理,首先dfs求得忽略无向边后的最长链点数k,那么问题就是判断是否可以通过无向边定向从而使得最长链点数不超过k。用dp的判断。
设f[i]表示以i为根的子树中所有的边定向后最长链点数不超过k条件下出边中最长链的最小值,g[i]表示以i为根的子树中所有的边定向后最长链点数不超过k条件下入边中最长链的最小值(最小值是最低限度,如果最小值都不可行那么问题不可行)。
两种情况(w为子节点):
- 与w的边中没有双向边:得出w.f_max与w.g_max判断与k的大小关系 如果超过k 回值INF否则回值w.f_max与w.g_max(max代表最长链)。没有选择的情况,已经定形。
- 与w的边中有双向边:定向双向边使满足k的限定下f与g尽量小。批量定向,求解f[u]的时候,将w按照f从小到大排序,依此枚举p,对于p将p之前的定向为出边,计算f[u]。同理求解g[u]。最后判断与k的关系。根据双向边定向选择最优结果。
代码:
// UVa1380 A Scheduling Problem
// Rujia Liu
#include<iostream>
#include<string>
#include<cstring>
#include<sstream>
#include<vector>
#include<algorithm>
using namespace std; const int maxn = + ;
const int INF = ; struct Edge {
int u, v, d; // d=1 means u->v, d=2 means v->u, d=0 means u-v
Edge(int u=, int v=, int d=):u(u),v(v),d(d){}
}; vector<Edge> edges[maxn];
int n, root, maxlen, f[maxn], g[maxn], have_father[maxn]; // maximal length of a DIRECTED path starting from u
int dfs(int u) {
int ans = ;
for(int i = ; i < edges[u].size(); i++) {
int v = edges[u][i].v;
if(edges[u][i].d == ) //u->v
ans = max(ans, dfs(v)+);
}
return ans;
} bool read_data() {
bool have_data = false;
int a, b;
n = ;
for(int i = ; i < maxn; i++) edges[i].clear();
memset(have_father, , sizeof(have_father)); while(cin >> a && a){
string str;
have_data = true;
if(a > n) n = a;
while(cin >> str && str != ""){
int len = str.length();
char dir = str[len-];
if(dir == 'd' || dir == 'u') str = str.substr(, len-);
stringstream ss(str);
ss >> b; // b is a's son
if(b > n) n = b;
have_father[b] = ;
if(dir == 'd'){
edges[a].push_back(Edge(a, b, )); // forward
edges[b].push_back(Edge(b, a, )); // backward
}else if(dir == 'u'){
edges[a].push_back(Edge(a, b, ));
edges[b].push_back(Edge(b, a, ));
}else{
edges[a].push_back(Edge(a, b, )); // it's a rooted tree, so we don't store edge to father
}
}
}
if(have_data) {
for(int i = ; i <= n; i++)
if(!have_father[i] && !edges[i].empty()) { root = i; break; }
}
return have_data;
} struct UndirectedSon {
int w, f, g;
UndirectedSon(int w=, int f=, int g=):w(w),f(f),g(g){}
}; bool cmp_f(const UndirectedSon& w1, const UndirectedSon& w2) {
return w1.f < w2.f;
} bool cmp_g(const UndirectedSon& w1, const UndirectedSon& w2) {
return w1.g < w2.g;
} // calculate f[i] and g[i]
// return true iff f[i] < INF
// f[i] is the minimal length of the longest "->u" path if all subtree paths have length <= maxlen
// g[i] is the minimal length of the longest "u->" path if all subtree paths have length <= maxlen
// f[i] = g[i] = INF if "all subtree paths have length <= maxlen" cannot be satisfied
bool dp(int i, int fa) {
if(edges[i].empty()) {
f[i] = g[i] = ;
return true;
}
vector<UndirectedSon> sons;
int f0 = , g0 = ; // f'[i] and g'[i] for directed sons // let f'[i] = max{f[w] | w->i}+1, g'[i] = max{g[w] | i->w}+1
// then we should change some undirected edges to ->u or u-> edges so that f'[i]+g'[i] <= maxlen
// then f[i] is the minimal f'[i] under this condition, and g[i] is the minimal g'[i]
for(int k = ; k < edges[i].size(); k++) {
int w = edges[i][k].v;
if(w == fa) continue; //ch != fa
dp(w, i); //Çó½âÍê×Ó½ÚµãºóÇó½âµ±Ç°½áµã
int d = edges[i][k].d;
if(d == ) sons.push_back(UndirectedSon(w, f[w], g[w]));
else if(d == ) g0 = max(g0, g[w]+);
else f0 = max(f0, f[w]+);
}
// If there is no undirected edges, we're done
if(sons.empty()) {
f[i] = f0; g[i] = g0;
if(f[i] + g[i] > maxlen) { f[i] = g[i] = INF; }
return f[i] < INF;
} f[i] = g[i] = INF; // to calculate f[i], we sort f[w] of undirected sons in increasing order and make first p edges to w->i
// then we calculate f'[i] and g'[i], check for f'[i]+g'[i] <= maxlen and update answer
int s = sons.size();
sort(sons.begin(), sons.end(), cmp_f);
int maxg[maxn]; // maxg[i] is max{sons[i].g, sons[i+1].g, ...}
maxg[s-] = sons[s-].g;
for(int k = s-; k >= ; k--)
maxg[k] = max(sons[k].g, maxg[k+]);
for(int p = ; p <= sons.size(); p++) {
int ff = f0, gg = g0;
if(p > ) ff = max(ff, sons[p-].f+);
if(p < sons.size()) gg = max(gg, maxg[p]+);
if(ff + gg <= maxlen) f[i] = min(f[i], ff);
} // g[i] is similar
sort(sons.begin(), sons.end(), cmp_g);
int maxf[maxn]; // maxf[i] is max{sons[i].f, sons[i+1].f, ...}
maxf[s-] = sons[s-].f;
for(int k = s-; k >= ; k--)
maxf[k] = max(sons[k].f, maxf[k+]);
for(int p = ; p <= sons.size(); p++) {
int ff = f0, gg = g0;
if(p > ) gg = max(gg, sons[p-].g+);
if(p < sons.size()) ff = max(ff, maxf[p]+);
if(ff + gg <= maxlen) g[i] = min(g[i], gg);
} return f[i] < INF;
} int main() {
while(read_data()) {
maxlen = ;
for(int i = ; i <= n; i++) maxlen = max(maxlen, dfs(i));
// Note: the problem asks for the number of nodes in path, but all the "lengths" above mean "number of edges"
if(dp(root, -)) cout << maxlen+ << "\n";
else cout << maxlen+ << "\n";
}
return ;
}
Code from Rujia
【暑假】[深入动态规划]UVa 1380 A Scheduling Problem的更多相关文章
- UVA 1380 A Scheduling Problem
题目链接:https://uva.onlinejudge.org/index.php?option=com_onlinejudge&Itemid=8&page=show_problem ...
- 【UVA 1380】 A Scheduling Problem (树形DP)
A Scheduling Problem Description There is a set of jobs, say x1, x2,..., xn <tex2html_verbatim_ ...
- UVa 101 The Blocks Problem Vector基本操作
UVa 101 The Blocks Problem 一道纯模拟题 The Problem The problem is to parse a series of commands that inst ...
- UVA - 524 Prime Ring Problem(dfs回溯法)
UVA - 524 Prime Ring Problem Time Limit:3000MS Memory Limit:0KB 64bit IO Format:%lld & % ...
- uva 10837 - A Research Problem(欧拉功能+暴力)
题目链接:uva 10837 - A Research Problem 题目大意:给定一个phin.要求一个最小的n.欧拉函数n等于phin 解题思路:欧拉函数性质有,p为素数的话有phip=p−1; ...
- UVA 810 - A Dicey Problem(BFS)
UVA 810 - A Dicey Problem 题目链接 题意:一个骰子,给你顶面和前面.在一个起点,每次能移动到周围4格,为-1,或顶面和该位置数字一样,那么问题来了,骰子能不能走一圈回到原地, ...
- UVA 10026 Shoemaker's Problem 鞋匠的难题 贪心+排序
题意:鞋匠一口气接到了不少生意,但是做鞋需要时间,鞋匠只能一双一双地做,根据协议每笔生意如果拖延了要罚钱. 给出每笔生意需要的天数和每天的罚钱数,求出最小罚钱的排列顺序. 只要按罚款/天数去从大到小排 ...
- UVA 1640 The Counting Problem UVA1640 求[a,b]或者[b,a]区间内0~9在里面各个数的数位上出现的总次数。
/** 题目:UVA 1640 The Counting Problem UVA1640 链接:https://vjudge.net/problem/UVA-1640 题意:求[a,b]或者[b,a] ...
- Uva 101 -- the block problem
Uva 101 the block problem 题目大意: 输入n,得到编号为0~n-1的木块,分别摆放在顺序排列编号为0~n-1的位置.现对这些木块进行操作,操作分为四种. 1.move a o ...
随机推荐
- 3.8 spring - AbstractBeanDefinition 介绍
之前,我们已尽完成了xml 文档到 GenericBeanDefinition的转换, 也就是说,到这里,所有的配置都可以在GenericBeanDefinition 的实例中找到对应的配置. Gen ...
- mybatis 的<![CDATA[ ]]>
示例: xml文件: <!-- 获取条数 --> <select id="getCount" parameterType="Map" resu ...
- PHP MSSQL数据操作PDO API
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 3 ...
- Spark 1.60的executor schedule
第一次看源码还是Spark 1.02.这次看新源码发现调度方式有了一些新的特征,在这里随便写一下. 不变的是,master还是接收Appclient和worker的消息,并且在接收RegisterAp ...
- 如何在CHROME里调试前端代码?
以前看前端们调得很神的, 刚看书到这里,作一个记录,演练了一下,确实有点神!!! :) <!DOCTYPE html> <html lang="en"> & ...
- HDU1565+状态压缩dp
简单的压缩状态 dp /* 状态压缩dp 同hdu2167 利用滚动数组!! */ #include<stdio.h> #include<string.h> #include& ...
- 压缩/解压 zip 时遇到 java.lang.IllegalArgumentException: MALFORMED
因为zip文件名为中文,或者压缩内容有中文 解决方法: 错误详情: SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinde ...
- redis其他问题
如何解决redis高并发客户端频繁time out? 现在业务上每天有5亿+的请求,平时redis的操作在2K+每秒左右.到了高峰有3K+,这时候客户端就会频繁的报connect time out的异 ...
- mybatis的知识点总结
1.接口绑定:两种方法,基于注解或者基于xml文档mapper,但要注意mapper的namespace要与接口路径完全一致. 2.orm格式转换:通过设置resultMap和ResultType,将 ...
- Java IO3:字符流
字符流 字节流提供了处理任何类型输入/输出操作的功能(对于计算机而言,一切都是0 和1,只需把数据以字节形式表示就够了),但它们不可以直接操作Unicode字符,一个Unicode字符占用2个字节,而 ...