#400712 (C++17) No.931 Multiplicative Convolution

• 問題: No.931 Multiplicative Convolution
• ユーザー: emthrm
• 提出日時: 2019-11-23 02:05:25
• 言語: C++17 (gcc 13.3.0 + boost 1.89.0)
• 結果: TLE (最新)
  AC
  (最初)
• 実行時間: -
• コード長: 10,434 bytes
• コンパイル時間: 1,783 ms
• コンパイル使用メモリ: 142,620 KB
• 最終ジャッジ日時: 2025-01-08 05:26:38
• ジャッジサーバーID （参考情報）: judge5 / judge2

ファイルパターン	結果
sample	AC * 3
other	AC * 13 TLE * 1

ソースコード

#include <algorithm>
#include <bitset>
#include <cassert>
#include <cctype>
#include <chrono>
#define _USE_MATH_DEFINES
#include <cmath>
#include <cstring>
#include <ctime>
#include <deque>
#include <functional>
#include <iostream>
#include <iomanip>
#include <iterator>
#include <map>
#include <numeric>
#include <queue>
#include <set>
#include <sstream>
#include <stack>
#include <string>
#include <tuple>
#include <utility>
#include <vector>
using namespace std;

#define FOR(i,m,n) for(int i=(m);i<(n);++i)
#define REP(i,n) FOR(i,0,n)
#define ALL(v) (v).begin(),(v).end()

const int INF = 0x3f3f3f3f;
const long long LINF = 0x3f3f3f3f3f3f3f3fLL;
const double EPS = 1e-8;
// const int MOD = 1000000007;
const int MOD = 998244353;
const int dy[] = {1, 0, -1, 0}, dx[] = {0, -1, 0, 1};
// const int dy[] = {1, 1, 0, -1, -1, -1, 0, 1},
//           dx[] = {0, -1, -1, -1, 0, 1, 1, 1};

struct IOSetup {
  IOSetup() {
    cin.tie(nullptr);
    ios_base::sync_with_stdio(false);
    cout << fixed << setprecision(20);
    cerr << fixed << setprecision(10);
  }
} iosetup;
/*-------------------------------------------------*/
int mod = MOD;
struct ModInt {
  unsigned val;
  ModInt(): val(0) {}
  ModInt(long long x) : val(x >= 0 ? x % mod : x % mod + mod) {}
  ModInt pow(long long exponent) {
    ModInt tmp = *this, res = 1;
    while (exponent > 0) {
      if (exponent & 1) res *= tmp;
      tmp *= tmp;
      exponent >>= 1;
    }
    return res;
  }
  ModInt &operator+=(const ModInt &rhs) { if((val += rhs.val) >= mod) val -= mod; return *this; }
  ModInt &operator-=(const ModInt &rhs) { if((val += mod - rhs.val) >= mod) val -= mod; return *this; }
  ModInt &operator*=(const ModInt &rhs) { val = static_cast<unsigned long long>(val) * rhs.val % mod; return *this; }
  ModInt &operator/=(const ModInt &rhs) { return *this *= rhs.inv(); }
  bool operator==(const ModInt &rhs) const { return val == rhs.val; }
  bool operator!=(const ModInt &rhs) const { return val != rhs.val; }
  bool operator<(const ModInt &rhs) const { return val < rhs.val; }
  bool operator<=(const ModInt &rhs) const { return val <= rhs.val; }
  bool operator>(const ModInt &rhs) const { return val > rhs.val; }
  bool operator>=(const ModInt &rhs) const { return val >= rhs.val; }
  ModInt operator-() const { return ModInt(val ? mod - val : 0); }
  ModInt operator+(const ModInt &rhs) const { return ModInt(*this) += rhs; }
  ModInt operator-(const ModInt &rhs) const { return ModInt(*this) -= rhs; }
  ModInt operator*(const ModInt &rhs) const { return ModInt(*this) *= rhs; }
  ModInt operator/(const ModInt &rhs) const { return ModInt(*this) /= rhs; }
  friend ostream &operator<<(ostream &os, const ModInt &rhs) { return os << rhs.val; }
  friend istream &operator>>(istream &is, ModInt &rhs) { long long x; is >> x; rhs = ModInt(x); return is; }
private:
  ModInt inv() const {
    // if (__gcd(val, mod) != 1) assert(false);
    unsigned a = val, b = mod; int x = 1, y = 0;
    while (b) {
      unsigned tmp = a / b;
      swap(a -= tmp * b, b);
      swap(x -= tmp * y, y);
    }
    return ModInt(x);
  }
};
int abs(const ModInt &x) { return x.val; }
struct Combinatorics {
  int val;
  vector<ModInt> fact, fact_inv, inv;
  Combinatorics(int val = 10000000) : val(val), fact(val + 1), fact_inv(val + 1), inv(val + 1) {
    fact[0] = 1;
    FOR(i, 1, val + 1) fact[i] = fact[i - 1] * i;
    fact_inv[val] = ModInt(1) / fact[val];
    for (int i = val; i > 0; --i) fact_inv[i - 1] = fact_inv[i] * i;
    FOR(i, 1, val + 1) inv[i] = fact[i - 1] * fact_inv[i];
  }
  ModInt nCk(int n, int k) {
    if (n < 0 || n < k || k < 0) return ModInt(0);
    // assert(n <= val && k <= val);
    return fact[n] * fact_inv[k] * fact_inv[n - k];
  }
  ModInt nPk(int n, int k) {
    if (n < 0 || n < k || k < 0) return ModInt(0);
    // assert(n <= val);
    return fact[n] * fact_inv[n - k];
  }
  ModInt nHk(int n, int k) {
    if (n < 0 || k < 0) return ModInt(0);
    return (k == 0 ? ModInt(1) : nCk(n + k - 1, k));
  }
};

namespace FFT {
  using Real = double;
  struct Complex {
    Real re, im;
    Complex(Real re = 0, Real im = 0) : re(re), im(im) {}
    inline Complex operator+(const Complex &rhs) const { return Complex(re + rhs.re, im + rhs.im); }
    inline Complex operator-(const Complex &rhs) const { return Complex(re - rhs.re, im - rhs.im); }
    inline Complex operator*(const Complex &rhs) const { return Complex(re * rhs.re - im * rhs.im, re * rhs.im + im * rhs.re); }
    inline Complex mul_real(Real r) const { return Complex(re * r, im * r); }
    inline Complex mul_pin(Real r) const { return Complex(-im * r, re * r); }
    inline Complex conj() const { return Complex(re, -im); }
  };

  vector<int> butterfly{0};
  vector<vector<Complex> > zeta{{{1, 0}}};

  void calc(int n) {
    int prev_n = butterfly.size();
    if (n <= prev_n) return;
    butterfly.resize(n);
    int prev_lg = zeta.size(), lg = __builtin_ctz(n);
    FOR(i, 1, prev_n) butterfly[i] <<= lg - prev_lg;
    FOR(i, prev_n, n) butterfly[i] = (butterfly[i >> 1] >> 1) | ((i & 1) << (lg - 1));
    zeta.resize(lg);
    FOR(i, prev_lg, lg) {
      zeta[i].resize(1 << i);
      Real angle = -M_PI * 2 / (1 << (i + 1));
      REP(j, 1 << (i - 1)) {
        zeta[i][j << 1] = zeta[i - 1][j];
        Real theta = angle * ((j << 1) + 1);
        zeta[i][(j << 1) + 1] = Complex(cos(theta), sin(theta));
      }
    }
  }

  void sub_dft(vector<Complex> &a) {
    int n = a.size();
    // assert(__builtin_popcount(n) == 1);
    calc(n);
    int shift = __builtin_ctz(butterfly.size()) - __builtin_ctz(n);
    REP(i, n) {
      int j = butterfly[i] >> shift;
      if (i < j) swap(a[i], a[j]);
    }
    for (int block = 1; block < n; block <<= 1) {
      int den = __builtin_ctz(block);
      for (int i = 0; i < n; i += (block << 1)) REP(j, block) {
        Complex tmp = a[i + j + block] * zeta[den][j];
        a[i + j + block] = a[i + j] - tmp;
        a[i + j] = a[i + j] + tmp;
      }
    }
  }

  template <typename T>
  vector<Complex> dft(const vector<T> &a) {
    int sz = a.size(), lg = 1;
    while ((1 << lg) < sz) ++lg;
    vector<Complex> c(1 << lg);
    REP(i, sz) c[i].re = a[i];
    sub_dft(c);
    return c;
  }

  vector<Real> real_idft(vector<Complex> &a) {
    int n = a.size(), half = n >> 1, quarter = half >> 1;
    // assert(__builtin_popcount(n) == 1);
    calc(n);
    a[0] = (a[0] + a[half] + (a[0] - a[half]).mul_pin(1)).mul_real(0.5);
    int den = __builtin_ctz(half);
    FOR(i, 1, quarter) {
      int j = half - i;
      Complex tmp1 = a[i] + a[j].conj(), tmp2 = ((a[i] - a[j].conj()) * zeta[den][j]).mul_pin(1);
      a[i] = (tmp1 - tmp2).mul_real(0.5);
      a[j] = (tmp1 + tmp2).mul_real(0.5).conj();
    }
    if (quarter > 0) a[quarter] = a[quarter].conj();
    a.resize(half);
    sub_dft(a);
    reverse(a.begin() + 1, a.end());
    Real r = 1.0 / half;
    vector<Real> res(n);
    REP(i, n) res[i] = (i & 1 ? a[i >> 1].im : a[i >> 1].re) * r;
    return res;
  }

  void idft(vector<Complex> &a) {
    int n = a.size();
    sub_dft(a);
    reverse(a.begin() + 1, a.end());
    Real r = 1.0 / n;
    REP(i, n) a[i] = a[i].mul_real(r);
  }

  template <typename T>
  vector<Real> convolution(const vector<T> &a, const vector<T> &b) {
    int a_sz = a.size(), b_sz = b.size(), sz = a_sz + b_sz - 1, lg = 1;
    while ((1 << lg) < sz) ++lg;
    int n = 1 << lg;
    vector<Complex> c(n);
    REP(i, a_sz) c[i].re = a[i];
    REP(i, b_sz) c[i].im = b[i];
    sub_dft(c);
    int half = n >> 1;
    c[0] = Complex(c[0].re * c[0].im, 0);
    FOR(i, 1, half) {
      Complex i_square = c[i] * c[i], j_square = c[n - i] * c[n - i];
      c[i] = (j_square.conj() - i_square).mul_pin(0.25);
      c[n - i] = (i_square.conj() - j_square).mul_pin(0.25);
    }
    c[half] = Complex(c[half].re * c[half].im, 0);
    vector<Real> res = real_idft(c);
    res.resize(sz);
    return res;
  }
};

vector<ModInt> convolution(const vector<ModInt> &a, const vector<ModInt> &b) {
  using Complex = FFT::Complex;
  const int pre = 15;
  int a_sz = a.size(), b_sz = b.size(), sz = a_sz + b_sz - 1, lg = 1;
  while ((1 << lg) < sz) ++lg;
  int n = 1 << lg;
  vector<Complex> A(n, 0), B(n, 0);
  REP(i, a_sz) {
    int ai = a[i].val;
    A[i] = Complex(ai & ((1 << pre) - 1), ai >> pre);
  }
  REP(i, b_sz) {
    int bi = b[i].val;
    B[i] = Complex(bi & ((1 << pre) - 1), bi >> pre);
  }
  FFT::sub_dft(A);
  FFT::sub_dft(B);
  int half = n >> 1;
  Complex tmp_a = A[0], tmp_b = B[0];
  A[0] = Complex(tmp_a.re * tmp_b.re, tmp_a.im * tmp_b.im);
  B[0] = Complex(tmp_a.re * tmp_b.im + tmp_a.im * tmp_b.re, 0);
  FOR(i, 1, half) {
    int j = n - i;
    Complex a_l_i = (A[i] + A[j].conj()).mul_real(0.5), a_h_i = (A[j].conj() - A[i]).mul_pin(0.5);
    Complex b_l_i = (B[i] + B[j].conj()).mul_real(0.5), b_h_i = (B[j].conj() - B[i]).mul_pin(0.5);
    Complex a_l_j = (A[j] + A[i].conj()).mul_real(0.5), a_h_j = (A[i].conj() - A[j]).mul_pin(0.5);
    Complex b_l_j = (B[j] + B[i].conj()).mul_real(0.5), b_h_j = (B[i].conj() - B[j]).mul_pin(0.5);
    A[i] = a_l_i * b_l_i + (a_h_i * b_h_i).mul_pin(1);
    B[i] = a_l_i * b_h_i + a_h_i * b_l_i;
    A[j] = a_l_j * b_l_j + (a_h_j * b_h_j).mul_pin(1);
    B[j] = a_l_j * b_h_j + a_h_j * b_l_j;
  }
  tmp_a = A[half]; tmp_b = B[half];
  A[half] = Complex(tmp_a.re * tmp_b.re, tmp_a.im * tmp_b.im);
  B[half] = Complex(tmp_a.re * tmp_b.im + tmp_a.im * tmp_b.re, 0);
  FFT::idft(A);
  FFT::idft(B);
  vector<ModInt> res(sz);
  ModInt tmp1 = 1 << pre, tmp2 = 1LL << (pre << 1);
  REP(i, sz) {
    res[i] += static_cast<long long>(A[i].re + 0.5);
    res[i] += tmp1 * static_cast<long long>(B[i].re + 0.5);
    res[i] += tmp2 * static_cast<long long>(A[i].im + 0.5);
  }
  return res;
}

int main() {
  int p; cin >> p;
  mod = p;
  int root = 2;
  while (true) {
    set<ModInt> st;
    FOR(i, 1, p) st.emplace(ModInt(root).pow(i));
    if (st.size() == p - 1) break;
    ++root;
  }
  vector<int> memo(p - 1);
  REP(i, p - 1) memo[i] = ModInt(root).pow(i).val;
  mod = MOD;
  vector<int> a(p, 0), b(p, 0);
  FOR(i, 1, p) cin >> a[i];
  FOR(i, 1, p) cin >> b[i];
  vector<ModInt> A(p - 1, 0), B(p - 1, 0);
  REP(i, p - 1) {
    A[i] = a[memo[i]];
    B[i] = b[memo[i]];
  }
  vector<ModInt> C = convolution(A, B);
  FOR(i, p - 1, C.size()) C[i % (p - 1)] += C[i];
  vector<ModInt> ans(p, 0);
  REP(i, p - 1) ans[memo[i]] = C[i];
  FOR(i, 1, p) cout << ans[i] << (i + 1 == p ? '\n' : ' ');
  return 0;
}