• 2017 Wuhan University Programming Contest (Online Round) E Lost in WHU(矩阵快速幂





    2017 Wuhan University Programming Contest (Online Round)



    Input file: standard input
    Output file: standard output
    Time limit: 1 second
    Memory limit: 512 mebibytes
    As one of the most beautiful campus in China, Wuhan University is around several hills, so the road is complex and visitors always lose themselves. Give a undirected graph of WHU of NN points and a deadline, then you should count the number of plan to reach the destination by deadline (the starting point is 1 and the ending point is NN).

    Input
    First line contains the number of points NN (Nle 100N≤100) and the number of edges MM (Mle N(N-1)/2M≤N(N−1)/2).

    The ii-th line of next MM lines contains two numbers u_iu
    ​i
    ​​ and v_iv
    ​i
    ​​ , which respents an edge (u_i, v_i)(u
    ​i
    ​​ ,v
    ​i
    ​​ ).

    The last line contains the deadline TT(Tle 10^9T≤10
    ​9
    ​​ ).

    Output
    The number of plan to reach the destination by deadline module 10^9+710
    ​9
    ​​ +7.

    Examples
    Input 1
    4 5
    1 3
    2 3
    3 4
    1 2
    1 4
    8
    Output 1
    170
    注意:数组开LL, mod = 1e9+7;
    题意:给你n个点,m条路,以及时间t,求在时间t内到达n点的方法总数;
    有两点需要理解。
    一:到达终点后不能继续走了,所以mt[n][i]=0;
    二:只有把mt[n][n]=1,求出来的才是在时间k内能走到终点的所有方法。
    举个例子
    比如说有这样一组数据
    1 4
    1 2
    2 3
    2 4
    3 4
    3

    然后构造初始矩阵A
    0 1 0 1
    1 0 1 1
    0 1 0 1
    0 0 0 1
    上面的mat[1][n]代表1到n需要走1单位时间的路径有几条

    然后是矩阵A^2
    1 0 1 2
    0 2 0 3
    1 0 1 2
    0 0 0 1
    这里第一行代表点1到点j需要走2单位时间的路径有几条

    最后一列代表点i到点n需要走<=2单位时间的路径有几条(这里就是令mat[n][n]=1的妙处了)

    矩阵A^3
    0 2 0 4
    2 0 2 5
    0 2 0 4
    0 0 0 1
    在矩阵A^3中mat[1][n]=4,所以答案就是4了(具体详见代码)
    ps:想一下矩阵是怎么乘的,就应该理解·了,我是这样理解的.(a[1][n]乘以a[n][n],才能加上小于k时间的方法,如果a[n][n]=0就加不上了,我是这样理解的)

    #include<stdio.h>
    #include<iostream>
    #include<string.h>
    #include<algorithm>
    using namespace std;
    #define mem(a,b) memset(a,b,sizeof(a))
    #define LL long long int 
    const int mod=1000000000+7;
    int n,m;
    struct mat
    {
        LL mt[105][105];
        mat()
        {
            memset(mt,0,sizeof(mt));
        }
    };
    mat I;
    mat cheng(mat q,mat p)
    {
        mat c;
        mem(c.mt,0);
        for(int i=0; i<n; i++)
            for(int j=0; j<n; j++)
                for(int k=0; k<n; k++)
                {
                    c.mt[i][j]=(c.mt[i][j]+q.mt[i][k]*p.mt[k][j])%mod;
                }
        return c;
    }
    int main()
    {
        while(~scanf("%d%d",&n,&m))
        {
    
            mat ans;
            mem(ans.mt,0);
            for(int i=0; i<m; i++)
            {
                int x,y;
                scanf("%d%d",&x,&y);
                if(x!=n) ans.mt[x-1][y-1]=1;//到达n后不能再走
                if(y!=n) ans.mt[y-1][x-1]=1;//到达n后不能再走
            }
            ans.mt[n-1][n-1]=1;//计算<=k步的路径总数,如果为0,则计算=k时的路径数  
            int k;
            scanf("%d",&k);
            I=ans;
            mem(ans.mt,0);
            for(int i=0; i<n; i++)
                ans.mt[i][i]=1;
            while(k)
            {
                if(k&1) ans=cheng(ans,I);
                I=cheng(I,I);
                k>>=1;
            }
            printf("%lld
    ",ans.mt[0][n-1]);
        }
        return 0;
    }
  • 相关阅读:
    神经网络回顾-感知机
    遗传算法杂记
    差分进化算法DE和粒子群算法PSO
    遗传算法GA
    Caffe学习 五 conv_layer与im2col
    Selenium
    Selenium
    Selenium
    Selenium
    Selenium
  • 原文地址:https://www.cnblogs.com/zxy160/p/7215118.html
Copyright © 2020-2023  润新知