hdu 2480 (贪心)

Steal the Treasure

Time Limit: 10000/6000 MS (Java/Others)    Memory Limit: 32768/32768 K (Java/Others)
Total Submission(s): 654    Accepted Submission(s): 162

Problem Description

The alliance of thieves decides to steal the treasure from country A. There are n cities in country A. Cities are connected by directional or bidirectional road. To avoid the risk, the king of country A divides his treasure and hides them in some place on the road.
  The alliance has found out the secret of the king. They get a map of country A which shows the location and the quantity of treasure on each road. In order to make the maximum profit and reduce the least loss, the alliance determines to send n thieves respectively to each city (one city one thief). At the appointed time, each thief chooses one road (if there is a road and notice that the road may have direction) to get to its corresponding city. Then he can steal the treasure on that road. After stealing, all the thieves return back to their base immediately.
  The heads of the alliance wonder to know the quantity of the treasure they can steal at most.

 

Input

There are multiple cases. Input is terminated by EOF.
  For each case, the first line contains two integers n (1<=n<=1000) and m (0<=m<=n*(n-1)/2), representing the number of cities and the number of roads in country A. The following m lines, each line contains four integers x, y (1<=x, y<=n, x≠y), d (0<=d<=1), w (0<=w<=1000), which means that there is a road from city x to city y, d=0 shows this road is bidirectional and d=1 shows it is directional and x the starting point, w is the quantity of treasure on the road.
  We guarantee that the road (x, y) and (y, x) will never appear together in the same case.

 

Output

For each case, output the maximum quantity of treasure the alliance can get.

 

Sample Input

  
  

   
   2 1
1 2 0 10
5 5
1 2 1 0
1 3 1 10
2 3 0 20
3 4 0 30
4 2 1 40
  
  

 

Sample Output

  
  

   
   10
100
  
  

 

Source

2008 Asia Regional Chengdu

 

Recommend

lcy

 

这题初看，很容易想到匹配的模型，用费用流去解，但看到数据范围，就知道这种做法是不行的。那怎么办，第一直觉是换策略，贪心怎么样？按边权的顺序排序后，从大到小去贪。对于有向边，只能选起点，那无向边呢，两个点都是可以的，而且怎么选择对后面是有影响的。于是想到先把这两个点标记为待确定点，根绝后面的选择再确定这两个点。注意到一个点可能是多条无向边的端点，也就是说可能有许多点互相关联，待确定，只要其中有一个点被确定了，那么其他所有的点就都被确定好了。这样这些点构成了一个联通块，可以用并查集实现。这样贪心的时候，每个联通块缩为一个点，每个点的状态有已被选，和未选，再根据相应的策略去贪心就好了。

#include <iostream>
#include <cstring>
#include <cstdio>
#include <string>
#include <algorithm>
#include <map>
#include <vector>
#include<queue>
#include<set>
using namespace std;
typedef long long LL;
typedef pair<int,int> P;
const int maxn = 1000 + 5;
const int INF = 1000000000;

struct Edge{
    int from,to,d,w;
}e[maxn*maxn];

bool cmp(Edge a,Edge b){
    return a.w < b.w;
}

int vis[maxn];
int fa[maxn];

int Find(int x){return x==fa[x]?x:fa[x]=Find(fa[x]);}

int main(){
    int n,m;
    while(scanf("%d%d",&n,&m) != EOF){
        for(int i = 0;i < m;i++){
            scanf("%d%d%d%d",&e[i].from,&e[i].to,&e[i].d,&e[i].w);
        }
        sort(e,e+m,cmp);
        memset(vis,0,sizeof(vis));
        for(int i = 0;i <= n;i++) fa[i] = i;
        int ans = 0;
        for(int i = m-1;i >= 0;i--){
            int x = Find(e[i].from);
            int y = Find(e[i].to);
            if(vis[x] && vis[y]) continue;
            else if(e[i].d == 1 && vis[x]) continue;
            ans += e[i].w;
            if(e[i].d == 1) vis[x] = 1;
            else{
                if(x == y) vis[x] = 1;
                if(vis[x] == 0 && vis[y] == 0 ) fa[x] = y;
                else if(vis[x] == 1) vis[y] = 1;
                else if(vis[y] == 1) vis[x] = 1;
            }
        }
        printf("%d\n",ans);
    }
    return 0;
}