zoukankan      html  css  js  c++  java
  • zhx's contest (矩阵快速幂 + 数学推论)

    zhx's contest

    Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 65536/65536 K (Java/Others) Total Submission(s): 851    Accepted Submission(s): 282

    Problem Description
    As one of the most powerful brushes, zhx is required to give his juniors n problems. zhx thinks the ith problem's difficulty is i. He wants to arrange these problems in a beautiful way. zhx defines a sequence {ai} beautiful if there is an i that matches two rules below: 1: a1..ai are monotone decreasing or monotone increasing. 2: ai..an are monotone decreasing or monotone increasing. He wants you to tell him that how many permutations of problems are there if the sequence of the problems' difficulty is beautiful. zhx knows that the answer may be very huge, and you only need to tell him the answer module p.
     
    Input
    Multiply test cases(less than 1000). Seek EOF as the end of the file. For each case, there are two integers n and p separated by a space in a line. (1n,p1018)
     
    Output
    For each test case, output a single line indicating the answer.
     
    Sample Input
    2 233 3 5
     
    Sample Output
    2 1
    Hint
    In the first case, both sequence {1, 2} and {2, 1} are legal. In the second case, sequence {1, 2, 3}, {1, 3, 2}, {2, 1, 3}, {2, 3, 1}, {3, 1, 2}, {3, 2, 1} are legal, so the answer is 6 mod 5 = 1
     
    Source
     
     1 #include<stdio.h>
     2 typedef long long ll ;
     3 ll n , mod ;
     4 
     5 ll mut (ll a , ll b)
     6 {
     7     ll ans = 0 ;
     8     while (b) {
     9         if (b & 1) {
    10             ans = (ans + a) % mod ;
    11         }
    12         a = (a + a ) % mod ;
    13         b >>= 1 ;
    14     }
    15     return ans % mod ;
    16 }
    17 
    18 void mgml (ll a , ll b)
    19 {
    20     ll ans = 1 ;
    21     while (b) {
    22         if (b & 1) {
    23             ans = mut (ans , a ) ;
    24         }
    25         b >>= 1 ;
    26         a = mut ( a , a ) ;
    27     }
    28     ans -= 2 ;
    29     ans %= mod ;
    30     ans += mod ;
    31     ans %= mod ;
    32     printf ("%I64d
    " , ans ) ;
    33 }
    34 
    35 int main ()
    36 {
    37     //freopen ("a.txt" , "r" , stdin ) ;
    38     while (~ scanf ("%I64d%I64d" , &n , &mod)) {
    39         if (n == 1) {
    40             printf ("%I64d
    " , 1LL % mod ) ;
    41             continue ;
    42         }
    43         mgml (2 , n ) ;
    44     }
    45     return 0 ;
    46 }
    View Code

    这道题到处是坑,推出2^n - 2照样进坑;
    首先要防止n = 1 , p = 1 这种情况;

    然后要防止ans - 2 < 0;

    然后  是 快速幂 的内部优化,让人耳目一新啊

  • 相关阅读:
    [转载] 美团-云鹏: 写给工程师的十条精进原则
    Docker测试一个静态网站
    Docker容器访问外部世界
    Docker容器间通信
    Docker网络(host、bridge、none)详细介绍
    Docker的资源限制(内存、CPU、IO)详细篇
    esxi中CentOS7不停机加磁盘并扩容现有分区
    ESXI6.5安装CentOS7教程
    Linux查看占用CPU和内存的 的程序
    Centos7使用脚本搭建LVS的DR模式。
  • 原文地址:https://www.cnblogs.com/get-an-AC-everyday/p/4340267.html
Copyright © 2011-2022 走看看