2018 Multi-University Training Contest 4

                **Problem D. Nothing is Impossible**
Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 262144/262144 K (Java/Others)
Total Submission(s): 145    Accepted Submission(s): 92


Problem Description
m students, including Kazari, will take an exam tomorrow.
The paper consists of exactly n problems, the i-th problem contains ai correct answers and bi incorrect answers, i.e. the i-th problem contains ai+bi candidates in total.
Each student should choose exactly one candidate as answer for each problem. If the answer to a certain problem is correct, then the student will get one point. The student who gets the most points wins.
Students only know the structure of the paper, but they are able to talk with each other during the exam. They decide to choose a subset S of all n problems, and they will only be able to submit answers on these problems.
They want to know the maximum size of S that the winner among them will solve all the problems in S if they take the optimal strategy.

For sample 1, students can choose S={1}，and we need at least 4 students to guarantee the winner solve the only problem.

For sample 2, students can choose S={1,2,3}, and we need at least 24 students to guarantee the winner solve these three problems, but if |S|=4, we need at least 96 students, which is more than 50.



Input
The first line of the input contains an integer T (1≤T≤100) denoting the number of test cases.
Each test case starts with two integers n,m (1≤n≤100,1≤m≤109), denoting the number of problems and the number of students. Each of next n lines contains two integers ai,bi （1≤bi≤100,ai=1）, indicating the number of correct answers and the number of incorrect answers of the i-th problem.


Output
For each test case, print an integer denoting the maximum size of S.



Sample Input
2   
3 5
1 3
1 3
1 3
5 50
1 1
1 3
1 2
1 3
1 5


Sample Output
1
3
题意：有n个问题 ，m个学生，每一个问题有ai个选项是正确的，bi个选项是错误的，共有ai+bi的选项，答对问题得一分，分最高的为获胜者，在采取最优策略的情况下，求获胜者至少问答问题正确的数目。
思路： 要使最优，则平分学生到每个选项，我们先对问题的选项数目从小到大排序，然后for循环m乘以问题的正确概率，到m为0是结束。

#include<string.h>  
#include<math.h>
#include<stdio.h>
#include<algorithm>
using namespace std;
typedef struct node {
    int a;
    int b;
}n;
n s[105];
int cmp(node x,node y)
{
    return (x.a+x.b)<(y.a +y.b);
}
int main()
{
     int t,n,m,i;
     scanf("%d",&t);
     while(t--)
     {
        scanf("%d%d",&n,&m);
        for( i=0;i<n;i++)
        {
            scanf("%d %d",&s[i].a,&s[i].b);
        }
        sort(s,s+n,cmp);
        for( i=0;i<n;i++)
        {
            m=m/(s[i].a+s[i].b);   
            //因为ai==0；所以概率为1/（s[i].a+s[i].b）.
            if(m==0)
            break;
        }
        printf("%d\n",i);
     }
     return 0;
}