• 【THUPC 2017】小L的计算题


    Problem

    Description

    现有一个长度为 (n) 的非负整数数组 ({a_i})。小 L 定义了一种神奇变换:

    [f_k=sum_{i=1}^{n}{a_i}^kpmod{ 998244353 } ]

    小 L 计划用变换生成的序列 (f) 做一些有趣的事情,但是他并不擅长算乘法,所以来找你帮忙,希望你能帮他尽快计算出 (f_1sim f_n)

    总共有 (T) 组数据。

    Range

    (nle 2 imes10^5, Tle20, sum nle 4 imes10^5, a_ile 10^9)

    Algorithm

    生成函数,多项式

    Mentality

    写出生成函数 (F) 的表达式:

    [F=sum_{k} f_kx^k\ =sum_{k} sum_{i=1}^n a_i^kx^k\ =sum_{i=1}^n sum_{k} (a_ix)^k\ =sum_{i=1}^n frac{1}{1-a_ix}\ =n-xsum_{i=1}^n frac{-a_i}{1-a_ix} ]

    然后发现 ((ln(1-a_ix))'=frac{-a_i}{1-a_ix}) ,直接代入:

    [F=n-xsum_{i=1}^n (ln(1-a_ix))'\ =n-x(sum_{i=1}^n ln(1-a_ix))'\ =n-x(ln(prod_{i=1}^n (1-a_ix)))' ]

    用分治计算 (prod) ,然后求个 (ln) 就完事了。

    Code

    #include <cmath>
    #include <cstdio>
    #include <iostream>
    #include <vector>
    using namespace std;
    #define LL long long
    #define go(G, x, i, v) 
      for (int i = G.hd[x], v = G.to[i]; i; v = G.to[i = G.nx[i]])
    #define inline __inline__ __attribute__((always_inline))
    inline LL read() {
      LL x = 0, w = 1;
      char ch = getchar();
      while (!isdigit(ch)) {
        if (ch == '-') w = -1;
        ch = getchar();
      }
      while (isdigit(ch)) {
        x = (x << 3) + (x << 1) + ch - '0';
        ch = getchar();
      }
      return x * w;
    }
    
    const int Max_n = 4e6 + 5, mod = 998244353;
    int T;
    bool fl;
    int n, cnt, a[Max_n];
    vector<int> f[1000000], ans;
    
    namespace Input {
    void main() {
      n = read();
      for (int i = 1; i <= n; i++) a[i] = read();
    }
    }  // namespace Input
    
    namespace Poly {
    int len, bit, rev[Max_n];
    int ksm(int a, int b = mod - 2) {
      int res = 1;
      for (; b; b >>= 1, a = (LL)a * a % mod)
        if (b & 1) res = (LL)res * a % mod;
      return res;
    }
    void init(int n) {
      len = 1 << (bit = log2(n) + 1);
      for (int i = 0; i < len; i++)
        rev[i] = rev[i >> 1] >> 1 | ((i & 1) << bit - 1);
    }
    void dft(vector<int> &f, bool t) {
      for (int i = 0; i < len; i++)
        if (rev[i] > i) swap(f[i], f[rev[i]]);
      for (int l = 1; l < len; l <<= 1) {
        int Wn = ksm(3, (mod - 1) / (l << 1));
        if (t) Wn = ksm(Wn);
        for (int i = 0; i < len; i += l << 1) {
          int Wnk = 1;
          for (int j = i; j < i + l; j++, Wnk = (LL)Wnk * Wn % mod) {
            int x = f[j], y = (LL)f[j + l] * Wnk % mod;
            f[j] = (x + y) % mod, f[j + l] = (x - y + mod) % mod;
          }
        }
      }
      if (t)
        for (int i = 0, Inv = ksm(len); i < len; i++) f[i] = (LL)f[i] * Inv % mod;
    }
    void Resize(vector<int> &f, int len) {
      f.resize(len);
      for (int i = 0; i < len; i++) f[i] = 0;
    }
    void Mul(vector<int> f, vector<int> &g, vector<int> &res, int N) {
      init(N);
      static vector<int> G;
      Resize(res, len), Resize(G, len);
      for (int i = 0; i < min((int)f.size(), len); i++) res[i] = f[i];
      for (int i = 0; i < min((int)g.size(), len); i++) G[i] = g[i];
      dft(res, 0), dft(G, 0);
      for (int i = 0; i < len; i++) res[i] = (LL)res[i] * G[i] % mod;
      dft(res, 1);
    }
    void Inv(vector<int> &f, vector<int> &res, int N) {
      init(N * 6);
      Resize(res, len);
      static vector<int> F;
      Resize(F, len);
      res[0] = ksm(f[0]);
      for (int deg = 2; deg < (N << 1); deg <<= 1) {
        init(deg * 3);
        for (int i = 0; i < min(deg, (int)f.size()); i++) F[i] = f[i];
        for (int i = min(deg, (int)f.size()); i < len; i++) F[i] = 0;
        dft(F, 0), dft(res, 0);
        for (int i = 0; i < len; i++)
          res[i] = (2ll * res[i] % mod + mod - (LL)res[i] * res[i] % mod * F[i] % mod) % mod;
        dft(res, 1);
        for (int i = deg; i < len; i++) res[i] = 0;
      }
    }
    void Ln(vector<int> &f, vector<int> &res, int N) {
      static vector<int> inv;
      res = f;
      for (int i = 0; i < N; i++) res[i] = (LL)res[i + 1] * (i + 1) % mod;
      res[N] = 0, Inv(f, inv, N);
      Mul(res, inv, res, N + N);
    }
    }  // namespace Poly
    using namespace Poly;
    
    namespace Solve {
    void Solve(int o, int l, int r) {
      if (l == r) {
        f[o].resize(2);
        f[o][0] = 1, f[o][1] = (-a[l] % mod + mod) % mod;
        return;
      }
      int mid = l + r >> 1;
      Solve(o << 1, l, mid), Solve(o << 1 | 1, mid + 1, r);
      Mul(f[o << 1], f[o << 1 | 1], f[o], r - l + 2);
    }
    void main() {
      Solve(1, 1, n);
      for (int i = n + 1; i < len; i++) f[1][i] = 0;
      fl = 1;
      Ln(f[1], ans, n + 1);
      int Ans = 0;
      for (int i = 0; i < n; i++) Ans ^= (-ans[i] + mod) % mod;
      cout << Ans << endl;
    }
    }  // namespace Solve
    
    int main() {
    #ifndef ONLINE_JUDGE
      freopen("2409.in", "r", stdin);
      freopen("2409.out", "w", stdout);
    #endif
      T = read();
      while (T--) {
        Input::main();
        Solve::main();
      }
    }
    
  • 相关阅读:
    Elasticsearch重要文章之四:监控每个节点(ThreadPool部分)
    [翻译]Elasticsearch重要文章之四:监控每个节点(jvm部分)
    IDFA
    Python中eval函数的作用
    Android开发重点难点:RelativeLayout(相对布局)详解
    布局相关
    Nginx配置proxy_pass
    Docker部署Elasticsearch集群
    extends Thread 与 implements Runnable 的区别
    玩转 Redis缓存 集群高可用
  • 原文地址:https://www.cnblogs.com/luoshuitianyi/p/12891662.html
Copyright © 2020-2023  润新知