#714203 (C) No.502 階乗を計算するだけ

提出ソース
結果

問題	No.502 階乗を計算するだけ
ユーザー	nonamae
提出日時	2021-11-08 22:49:06
言語	C (gcc 12.3.0)
結果	WA
実行時間	-
コード長	20,461 bytes
コンパイル時間	2,607 ms
コンパイル使用メモリ	74,632 KB
実行使用メモリ	17,540 KB
最終ジャッジ日時	2024-04-28 10:38:27
合計ジャッジ時間	9,385 ms
ジャッジサーバーID （参考情報）	judge5 / judge2
このコードへのチャレンジ
（要ログイン）
テストケース

テストケース表示
入力	結果	実行時間実行使用メモリ
testcase_00	AC	10 ms 12,680 KB
testcase_01	AC	11 ms 12,712 KB
testcase_02	AC	138 ms 17,168 KB
testcase_03	AC	138 ms 16,484 KB
testcase_04	AC	138 ms 16,224 KB
testcase_05	AC	138 ms 15,588 KB
testcase_06	AC	139 ms 15,508 KB
testcase_07	AC	139 ms 15,592 KB
testcase_08	AC	138 ms 15,648 KB
testcase_09	AC	139 ms 17,040 KB
testcase_10	AC	139 ms 15,572 KB
testcase_11	AC	138 ms 15,936 KB
testcase_12	AC	138 ms 15,640 KB
testcase_13	AC	138 ms 16,588 KB
testcase_14	AC	139 ms 16,824 KB
testcase_15	AC	138 ms 16,492 KB
testcase_16	AC	140 ms 16,592 KB
testcase_17	AC	138 ms 16,852 KB
testcase_18	AC	138 ms 15,576 KB
testcase_19	AC	139 ms 17,196 KB
testcase_20	AC	139 ms 16,260 KB
testcase_21	AC	139 ms 16,752 KB
testcase_22	WA	-
testcase_23	WA	-
testcase_24	WA	-
testcase_25	WA	-
testcase_26	WA	-
testcase_27	WA	-
testcase_28	WA	-
testcase_29	WA	-
testcase_30	WA	-
testcase_31	WA	-
testcase_32	WA	-
testcase_33	WA	-
testcase_34	WA	-
testcase_35	WA	-
testcase_36	WA	-
testcase_37	WA	-
testcase_38	WA	-
testcase_39	WA	-
testcase_40	WA	-
testcase_41	WA	-
testcase_42	AC	10 ms 12,216 KB
testcase_43	AC	11 ms 12,332 KB
testcase_44	AC	11 ms 13,636 KB
testcase_45	AC	11 ms 12,436 KB
testcase_46	AC	10 ms 12,156 KB
testcase_47	AC	10 ms 13,248 KB
testcase_48	AC	10 ms 13,676 KB
testcase_49	AC	11 ms 12,072 KB
testcase_50	AC	10 ms 13,304 KB
testcase_51	AC	11 ms 13,280 KB
権限があれば一括ダウンロードができます
ソースコード

raw source code
#pragma region opt
#pragma GCC target("avx2")
#pragma GCC optimize("O3")
#pragma endregion opt

#pragma region header
#define _GNU_SOURCE
#include <stdbool.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <assert.h>
#include <limits.h>
#include <math.h>
#include <string.h>
#include <time.h>
#pragma endregion header

#pragma region type
/* signed integer */
typedef   int8_t      i8;
typedef   int16_t     i16;
typedef   int32_t     i32;
typedef   int64_t     i64;
typedef __int128_t    i128;
/* unsigned integer */
typedef   uint8_t     u8;
typedef   uint16_t    u16;
typedef   uint32_t    u32;
typedef   uint64_t    u64;
typedef __uint128_t   u128;
/* floating point number */
typedef   float       f32;
typedef   double      f64;
typedef   long double f80;
#pragma endregion type

#pragma region macro
#define MIN(a, b) (((a) < (b)) ? (a) : (b))
#define MAX(a, b) (((a) > (b)) ? (a) : (b))
#define SWAP(a, b) (((a) ^= (b)), ((b) ^= (a)), ((a) ^= (b)))
#define POPCNT32(a) __builtin_popcount((a))
#define POPCNT64(a) __builtin_popcountll((a))
#define CTZ32(a) __builtin_ctz((a))
#define CLZ32(a) __builtin_clz((a))
#define CTZ64(a) __builtin_ctzll((a))
#define CLZ64(a) __builtin_clzll((a))
#define HAS_SINGLE_BIT32(a) (__builtin_popcount((a)) == (1))
#define HAS_SINGLE_BIT64(a) (__builtin_popcountll((a)) == (1))
#define MSB32(a) ((31) - __builtin_clz((a)))
#define MSB64(a) ((63) - __builtin_clzll((a)))
#define BIT_WIDTH32(a) ((a) ? ((32) - __builtin_clz((a))) : (0))
#define BIT_WIDTH64(a) ((a) ? ((64) - __builtin_clzll((a))) : (0))
#define LSBit(a) ((a) & (-(a)))
#define CLSBit(a) ((a) & ((a) - (1)))
#define BIT_CEIL32(a) ((!(a)) ? (1) : ((POPCNT32(a)) == (1) ? ((1u) << ((31) - CLZ32((a)))) : ((1u) << ((32) - CLZ32(a)))))
#define BIT_CEIL64(a) ((!(a)) ? (1) : ((POPCNT64(a)) == (1) ? ((1ull) << ((63) - CLZ64((a)))) : ((1ull) << ((64) - CLZ64(a)))))
#define BIT_FLOOR32(a) ((!(a)) ? (0) : ((1u) << ((31) - CLZ32((a)))))
#define BIT_FLOOR64(a) ((!(a)) ? (0) : ((1ull) << ((63) - CLZ64((a)))))
#define _ROTL32(x, s) (((x) << ((s) % (32))) | (((x) >> ((32) - ((s) % (32))))))
#define _ROTR32(x, s) (((x) >> ((s) % (32))) | (((x) << ((32) - ((s) % (32))))))
#define ROTL32(x, s) (((s) == (0)) ? (x) : ((((i64)(s)) < (0)) ? (_ROTR32((x), -(s))) : (_ROTL32((x), (s)))))
#define ROTR32(x, s) (((s) == (0)) ? (x) : ((((i64)(s)) < (0)) ? (_ROTL32((x), -(s))) : (_ROTR32((x), (s)))))
#define _ROTL64(x, s) (((x) << ((s) % (64))) | (((x) >> ((64) - ((s) % (64))))))
#define _ROTR64(x, s) (((x) >> ((s) % (64))) | (((x) << ((64) - ((s) % (64))))))
#define ROTL64(x, s) (((s) == (0)) ? (x) : ((((i128)(s)) < (0)) ? (_ROTR64((x), -(s))) : (_ROTL64((x), (s)))))
#define ROTR64(x, s) (((s) == (0)) ? (x) : ((((i128)(s)) < (0)) ? (_ROTL64((x), -(s))) : (_ROTR64((x), (s)))))
#pragma endregion macro

#pragma region io
int read_int(void) {
  // -2147483648 ～ 2147483647 (> 10 ^ 9)
  int c, x = 0, f = 1;
  while (c = getchar_unlocked(), c < 48 || c > 57) if (c == 45) f = -f;
  while (47 < c && c < 58) {
    x = x * 10 + c - 48;
    c = getchar_unlocked();
  }
  return f * x;
}
i32 in_i32(void) {
  // -2147483648 ～ 2147483647 (> 10 ^ 9)
  i32 c, x = 0, f = 1;
  while (c = getchar_unlocked(), c < 48 || c > 57) if (c == 45) f = -f;
  while (47 < c && c < 58) {
    x = x * 10 + c - 48;
    c = getchar_unlocked();
  }
  return f * x;
}
u32 in_u32(void) {
  // 0 ～ 4294967295 (> 10 ^ 9)
  u32 c, x = 0;
  while (c = getchar_unlocked(), c < 48 || c > 57);
  while (47 < c && c < 58) {
    x = x * 10 + c - 48;
    c = getchar_unlocked();
  }
  return x;
}
i64 in_i64(void) {
  // -9223372036854775808 ～ 9223372036854775807 (> 10 ^ 18)
  i64 c, x = 0, f = 1;
  while (c = getchar_unlocked(), c < 48 || c > 57) if (c == 45) f = -f;
  while (47 < c && c < 58) {
    x = x * 10 + c - 48;
    c = getchar_unlocked();
  }
  return f * x;
}
u64 in_u64(void) {
  // 0 ～ 18446744073709551615 (> 10 ^ 19)
  u64 c, x = 0;
  while (c = getchar_unlocked(), c < 48 || c > 57);
  while (47 < c && c < 58) {
    x = x * 10 + c - 48;
    c = getchar_unlocked();
  }
  return x;
}
static inline void write_int_inner(int x) {
  if (x >= 10) write_int_inner(x / 10);
  putchar_unlocked(x - x / 10 * 10 + 48);
}
void write_int(int x) {
  if (x < 0) {
    putchar_unlocked('-');
    x = -x;
  }
  write_int_inner(x);
}
static inline void out_i32_inner(i32 x) {
  if (x >= 10) out_i32_inner(x / 10);
  putchar_unlocked(x - x / 10 * 10 + 48);
}
void out_i32(i32 x) {
  if (x < 0) {
    putchar_unlocked('-');
    x = -x;
  }
  out_i32_inner(x);
}
void out_u32(u32 x) {
  if (x >= 10) out_u32(x / 10);
  putchar_unlocked(x - x / 10 * 10 + 48);
}
static inline void out_i64_inner(i64 x) {
  if (x >= 10) out_i64_inner(x / 10);
  putchar_unlocked(x - x / 10 * 10 + 48);
}
void out_i64(i64 x) {
  if (x < 0) {
    putchar_unlocked('-');
    x = -x;
  }
  out_i64_inner(x);
}
void out_u64(u64 x) {
  if (x >= 10) out_u64(x / 10);
  putchar_unlocked(x - x / 10 * 10 + 48);
}
void NL(void) { putchar_unlocked('\n'); }
void SP(void) { putchar_unlocked(' '); }
void write_int_array(int *a, int a_len) {
  for (int i = 0; i < a_len; i++) {
    if (i) SP();
    write_int(a[i]);
  }
  NL();
}
void out_i32_array(i32 *a, int a_len) {
  for (int i = 0; i < a_len; i++) {
    if (i) SP();
    out_i32(a[i]);
  }
  NL();
}
void out_u32_array(u32 *a, int a_len) {
  for (int i = 0; i < a_len; i++) {
    if (i) SP();
    out_u32(a[i]);
  }
  NL();
}
void out_i64_array(i64 *a, int a_len) {
  for (int i = 0; i < a_len; i++) {
    if (i) SP();
    out_i64(a[i]);
  }
  NL();
}
void out_u64_array(u64 *a, int a_len) {
  for (int i = 0; i < a_len; i++) {
    if (i) SP();
    out_u64(a[i]);
  }
  NL();
}
#pragma endregion io

#pragma region m32
typedef uint32_t m32;
m32 _one_m32(u32 mod) { return (u32)-1u % mod + 1; }
m32 _r2_m32(u32 mod) { return (u64)(i64)-1 % mod + 1; }
m32 _inv_m32(u32 mod) {
  u32 inv = mod;
  for (int i = 0; i < 4; ++i) inv *= 2 - inv * mod;
  return inv;
/**
  u32 u = 1, v = 0, x = 1u << 31;
  for (int i = 0; i < 32; i++) {
    if (u & 1) u = (u + mod) >> 1, v = (v >> 1) + x;
    else u >>= 1, v >>= 1;
  }
  return -v;
*/
}
m32 _reduce_m32(u64 a, m32 inv, u32 mod) {
  u32 y = (u32)(a >> 32) - (u32)(((u64)((u32)a * inv) * mod) >> 32);
  return (i32)y < 0 ? y + mod : y;
}
m32 to_m32(u32 a, m32 r2, m32 inv, u32 mod) { return _reduce_m32((u64)a * r2, inv, mod); }
u32 from_m32(m32 A, m32 inv, u32 mod) { return _reduce_m32(A, inv, mod); }
m32 add_m32(m32 A, m32 B, u32 mod) {
  return A + B >= mod ? A + B - mod:  A + B;
}
m32 sub_m32(m32 A, m32 B, u32 mod) {
  return A >= B ? A - B : mod + A - B;
}
m32 min_m32(m32 A, u32 mod) { return sub_m32(0u, A, mod); }
m32 mul_m32(m32 A, m32 B, m32 inv, u32 mod) { return _reduce_m32((u64)A * B, inv, mod); }
m32 pow_m32(m32 A, i32 n, m32 inv, u32 mod) {
  m32 ret = _one_m32(mod);
  while (n > 0) {
    if (n & 1) ret = mul_m32(ret, A, inv, mod);
    A = mul_m32(A, A, inv, mod);
    n >>= 1;
  }
  return ret;
}
m32 inv_m32(m32 A, m32 inv, u32 mod) { return pow_m32(A, (i32)mod - 2, inv, mod); }
m32 div_m32(m32 A, m32 B, m32 inv, u32 mod) {
  /* assert(is_prime(mod)); */
  return mul_m32(A, inv_m32(B, inv, mod), inv, mod);
}
m32 in_m32(m32 r2, m32 inv, u32 mod) {
  u32 c, a = 0;
  while (c = getchar_unlocked(), c < 48 || c > 57);
  while (47 < c && c < 58) {
    a = a * 10 + c - 48;
    c = getchar_unlocked();
  }
  return to_m32(a, r2, inv, mod);
}
void out_m32(m32 A, m32 inv, u32 mod) {
  u32 a = from_m32(A, inv, mod);
  out_u32(a);
}
#pragma endregion m32

#pragma region ntt

#pragma region ntt1
const u32 m1       = 998244353u;
const m32 m1_r2    = 932051910u;
const m32 m1_inv   = 3296722945u;
const m32 m1_one   = 301989884u;
const m32 m1_rev   = 696254469u;
const m32 m1_gs[]  = { 691295370, 307583142, 566821959, 878217029, 375146819, 138254384, 500602490, 79119218,  790898700, 978335284, 651424567, 308706579, 723000027, 474797508, 683394121, 44141573, 536892010, 945865189, 175417726, 536169764, 831722880, 721458245 };
const m32 m1_igs[] = { 306948983, 888603487, 138723248, 65668869,  842568658, 953245971, 195169681, 118717521, 792052763, 828450244, 908724728, 218560432, 628507989, 248210924, 566568154, 6285593,  82571768,  49985074,  225413092, 349167278, 61514562,  763211248 };
void ntt1(m32 *A, int A_len) {
  int h = 0;
  while (A_len > (1 << h)) h++;
  for (int ph = 1; ph <= h; ph++) {
    int w = 1 << (ph - 1);
    int p = 1 << (h - ph);
    m32 now = m1_one;
    for (int s = 0; s < w; s++) {
      int offset = s << (h - ph + 1);
      for (int i = 0; i < p; i++) {
        m32 l = A[i + offset];
        m32 r = mul_m32(A[i + offset + p], now, m1_inv, m1);
        A[i + offset] = add_m32(l, r, m1);
        A[i + offset + p] = sub_m32(l, r, m1);
      }
      now = mul_m32(now, m1_gs[CTZ32(~s)], m1_inv, m1);
    }
  }
}
void intt1(m32 *A, int A_len) {
  int h = 0;
  while (A_len > (1 << h)) h++;
  for (int ph = h; ph >= 1; ph--) {
    int w = 1 << (ph - 1);
    int p = 1 << (h - ph);
    m32 inow = m1_one;
    for (int s = 0; s < w; s++) {
      int offset = s << (h - ph + 1);
      for (int i = 0; i < p; i++) {
        m32 l = A[i + offset];
        m32 r = A[i + offset + p];
        A[i + offset] = add_m32(l, r, m1);
        A[i + offset + p] = mul_m32(sub_m32(l, r, m1), inow, m1_inv, m1);
      }
      inow = mul_m32(inow, m1_igs[CTZ32(~s)], m1_inv, m1);
    }
  }
  m32 inv2t = inv_m32(to_m32(A_len, m1_r2, m1_inv, m1), m1_inv, m1);
  for (int i = 0; i < A_len; i++) A[i] = mul_m32(A[i], inv2t, m1_inv, m1);
}
m32 *convolute1(m32 *A, int A_len, m32 *B, int B_len) {
  int ret_len = BIT_CEIL32(A_len + B_len - 1);
  m32 *C = (m32 *)calloc(ret_len, sizeof(m32));
  m32 *D = (m32 *)calloc(ret_len, sizeof(m32));
  #ifdef LOCAL
  if (C == NULL || D == NULL) exit(EXIT_FAILURE);
  #endif
  memcpy(C, A, sizeof(m32) * A_len);
  memcpy(D, B, sizeof(m32) * B_len);
  ntt1(C, ret_len);
  ntt1(D, ret_len);
  for (int i = 0 ; i < ret_len; i++) C[i] = mul_m32(C[i], D[i], m1_inv, m1);
  free(D);
  intt1(C, ret_len);
  return C;
}
#pragma endregion ntt1

#pragma region ntt2
const u32 m2       = 985661441u;
const m32 m2_r2    = 616455619u;
const m32 m2_inv   = 3309305857u;
const m32 m2_one   = 352321532u;
const m32 m2_rev   = 633339909u;
const m32 m2_gs[]  = { 969414155, 240156868, 716651500, 728800531, 977177032, 47314842,  240475723, 876076444, 626710676, 365360170, 808202916, 560909592, 755542104, 303317332, 75348256,  259192271, 882296372, 620044766, 876870197, 256206930, 761331788 };
const m32 m2_igs[] = { 16247286,  67104299,  325946810, 44505332,  582782266, 729124870, 724673072, 173952869, 594582867, 76943556,  66752559,  892797276, 469283465, 123325105, 933929770, 911329874, 741246559, 905930185, 828158135, 9523962,   198022420 };
void ntt2(m32 *A, int A_len) {
  int h = 0;
  while (A_len > (1 << h)) h++;
  for (int ph = 1; ph <= h; ph++) {
    int w = 1 << (ph - 1);
    int p = 1 << (h - ph);
    m32 now = m2_one;
    for (int s = 0; s < w; s++) {
      int offset = s << (h - ph + 1);
      for (int i = 0; i < p; i++) {
        m32 l = A[i + offset];
        m32 r = mul_m32(A[i + offset + p], now, m2_inv, m2);
        A[i + offset] = add_m32(l, r, m2);
        A[i + offset + p] = sub_m32(l, r, m2);
      }
      now = mul_m32(now, m2_gs[CTZ32(~s)], m2_inv, m2);
    }
  }
}
void intt2(m32 *A, int A_len) {
  int h = 0;
  while (A_len > (1 << h)) h++;
  for (int ph = h; ph >= 1; ph--) {
    int w = 1 << (ph - 1);
    int p = 1 << (h - ph);
    m32 inow = m2_one;
    for (int s = 0; s < w; s++) {
      int offset = s << (h - ph + 1);
      for (int i = 0; i < p; i++) {
        m32 l = A[i + offset];
        m32 r = A[i + offset + p];
        A[i + offset] = add_m32(l, r, m2);
        A[i + offset + p] = mul_m32(sub_m32(l, r, m2), inow, m2_inv, m2);
      }
      inow = mul_m32(inow, m2_igs[CTZ32(~s)], m2_inv, m2);
    }
  }
  m32 inv2t = inv_m32(to_m32(A_len, m2_r2, m2_inv, m2), m2_inv, m2);
  for (int i = 0; i < A_len; i++) A[i] = mul_m32(A[i], inv2t, m2_inv, m2);
}
m32 *convolute2(m32 *A, int A_len, m32 *B, int B_len) {
  int ret_len = BIT_CEIL32(A_len + B_len - 1);
  m32 *C = (m32 *)calloc(ret_len, sizeof(m32));
  m32 *D = (m32 *)calloc(ret_len, sizeof(m32));
  #ifdef LOCAL
  if (C == NULL || D == NULL) exit(EXIT_FAILURE);
  #endif
  memcpy(C, A, sizeof(m32) * A_len);
  memcpy(D, B, sizeof(m32) * B_len);
  ntt2(C, ret_len);
  ntt2(D, ret_len);
  for (int i = 0 ; i < ret_len; i++) C[i] = mul_m32(C[i], D[i], m2_inv, m2);
  free(D);
  intt2(C, ret_len);
  return C;
}
#pragma endregion ntt2

#pragma region ntt3
const u32 m3       = 943718401u;
const m32 m3_r2    = 917135855u;
const m32 m3_inv   = 3351248897u;
const m32 m3_one   = 520093692u;
const m32 m3_rev   = 423624709u;
const m32 m3_gs[]  = { 125689310, 401270432, 193546243, 204233475, 765072983, 793690592, 598110941, 560814539, 323055569, 635997590, 661263945, 671645950, 596439462, 577210208, 667936112, 172603057, 698142776, 3390265,   400541812, 419143563, 100582761 };
const m32 m3_igs[] = { 818029091, 177917178, 278610320, 675646939, 629165784, 803573782, 552038920, 685763768, 343497720, 610893888, 604907871, 366961343, 132493990, 882172703, 730481417, 529389095, 864269596, 777879390, 446333578, 468025435, 879098724 };
void ntt3(m32 *A, int A_len) {
  int h = 0;
  while (A_len > (1 << h)) h++;
  for (int ph = 1; ph <= h; ph++) {
    int w = 1 << (ph - 1);
    int p = 1 << (h - ph);
    m32 now = m3_one;
    for (int s = 0; s < w; s++) {
      int offset = s << (h - ph + 1);
      for (int i = 0; i < p; i++) {
        m32 l = A[i + offset];
        m32 r = mul_m32(A[i + offset + p], now, m3_inv, m3);
        A[i + offset] = add_m32(l, r, m3);
        A[i + offset + p] = sub_m32(l, r, m3);
      }
      now = mul_m32(now, m3_gs[CTZ32(~s)], m3_inv, m3);
    }
  }
}
void intt3(m32 *A, int A_len) {
  int h = 0;
  while (A_len > (1 << h)) h++;
  for (int ph = h; ph >= 1; ph--) {
    int w = 1 << (ph - 1);
    int p = 1 << (h - ph);
    m32 inow = m3_one;
    for (int s = 0; s < w; s++) {
      int offset = s << (h - ph + 1);
      for (int i = 0; i < p; i++) {
        m32 l = A[i + offset];
        m32 r = A[i + offset + p];
        A[i + offset] = add_m32(l, r, m3);
        A[i + offset + p] = mul_m32(sub_m32(l, r, m3), inow, m3_inv, m3);
      }
      inow = mul_m32(inow, m3_igs[CTZ32(~s)], m3_inv, m3);
    }
  }
  m32 inv2t = inv_m32(to_m32(A_len, m3_r2, m3_inv, m3), m3_inv, m3);
  for (int i = 0; i < A_len; i++) A[i] = mul_m32(A[i], inv2t, m3_inv, m3);
}
m32 *convolute3(m32 *A, int A_len, m32 *B, int B_len) {
  int ret_len = BIT_CEIL32(A_len + B_len - 1);
  m32 *C = (m32 *)calloc(ret_len, sizeof(m32));
  m32 *D = (m32 *)calloc(ret_len, sizeof(m32));
  #ifdef LOCAL
  if (C == NULL || D == NULL) exit(EXIT_FAILURE);
  #endif
  memcpy(C, A, sizeof(m32) * A_len);
  memcpy(D, B, sizeof(m32) * B_len);
  ntt3(C, ret_len);
  ntt3(D, ret_len);
  for (int i = 0 ; i < ret_len; i++) C[i] = mul_m32(C[i], D[i], m3_inv, m3);
  free(D);
  intt3(C, ret_len);
  return C;
}
#pragma endregion ntt3

#pragma endregion ntt

#pragma region convolute mod1000000007
const u32 mod = 1000000007u;
const m32 r2  = 582344008u;
const m32 inv = 2068349879u;
const m32 one = 294967268u;
u32 *convolute_mod1000000007(u32 *a, int a_len, u32 *b, int b_len) {
  m32 *A = (m32 *)calloc(a_len, sizeof(m32));
  m32 *B = (m32 *)calloc(b_len, sizeof(m32));
  #ifdef LOCAL
  if (A == NULL || B == NULL) exit(EXIT_FAILURE);
  #endif
  for (int i = 0; i < a_len; i++) A[i] = to_m32(a[i] % m1, m1_r2, m1_inv, m1);
  for (int i = 0; i < b_len; i++) B[i] = to_m32(b[i] % m1, m1_r2, m1_inv, m1);
  m32 *C1 = convolute1(A, a_len, B, b_len);
  for (int i = 0; i < a_len; i++) A[i] = to_m32(a[i] % m2, m2_r2, m2_inv, m2);
  for (int i = 0; i < b_len; i++) B[i] = to_m32(b[i] % m2, m2_r2, m2_inv, m2);
  m32 *C2 = convolute2(A, a_len, B, b_len);
  for (int i = 0; i < a_len; i++) A[i] = to_m32(a[i] % m3, m3_r2, m3_inv, m3);
  for (int i = 0; i < b_len; i++) B[i] = to_m32(b[i] % m3, m3_r2, m3_inv, m3);
  m32 *C3 = convolute3(A, a_len, B, b_len);
  free(A);
  free(B);
  u32 *ret = (u32 *)calloc(a_len + b_len - 1, sizeof(u32));
  #ifdef LOCAL
  if (ret == NULL) exit(EXIT_FAILURE);
  #endif
  m32 m1_inv_m2 = inv_m32(to_m32(m1, m2_r2, m2_inv, m2), m2_inv, m2);
  m32 m12_inv_m3 = inv_m32(mul_m32(to_m32(m1, m3_r2, m3_inv, m3), to_m32(m2, m3_r2, m3_inv, m3), m3_inv, m3), m3_inv, m3);
  m32 m1_m3 = to_m32(m1, m3_r2, m3_inv, m3);
  m32 m1_m0 = to_m32(m1, r2, inv, 1000000007u);
  m32 m12_m0 = mul_m32(to_m32(m1, r2, inv, 1000000007u), to_m32(m2, r2, inv, 1000000007u), inv, 1000000007u);
  for (int i = 0; i < a_len + b_len - 1; ++i) {
    u32 xi = from_m32(C1[i], m1_inv, m1);
    u32 yi = from_m32(C2[i], m2_inv, m2);
    u32 zi = from_m32(C3[i], m3_inv, m3);
    m32 xi_m2 = to_m32(xi, m2_r2, m2_inv, m2);
    m32 yi_m2 = to_m32(yi, m2_r2, m2_inv, m2);
    m32 zi_m3 = to_m32(zi, m3_r2, m3_inv, m3);
    m32 xi_m3 = to_m32(xi, m3_r2, m3_inv, m3);
    u32 v1 = from_m32(mul_m32(sub_m32(yi_m2, xi_m2, m2), m1_inv_m2, m2_inv, m2), m2_inv, m2);
    m32 v1_m3 = to_m32(v1, m3_r2, m3_inv, m3);
    u32 v2 = from_m32(mul_m32(sub_m32(zi_m3, add_m32(xi_m3, mul_m32(m1_m3, v1_m3, m3_inv, m3), m3), m3), m12_inv_m3, m3_inv, m3), m3_inv, m3);
    m32 v2_m0 = to_m32(v2, r2, inv, 1000000007u);
    m32 xi_m0 = to_m32(xi, r2, inv, 1000000007u);
    m32 v1_m0 = to_m32(v1, r2, inv, 1000000007u);
    ret[i] = from_m32(add_m32(add_m32(xi_m0, mul_m32(m1_m0, v1_m0, inv, 1000000007u), 1000000007u), mul_m32(m12_m0, v2_m0, inv, 1000000007u), 1000000007u), inv, 1000000007u);
  }
  free(C1);
  free(C2);
  free(C3);
  return ret;
}
#pragma endregion convolute mod1000000007

#pragma region sample_point_shift
m32 _fact[1<<20];
m32 _inv_fact[1<<20];
m32 _inv_table[1<<20];
void pre_fact(int n) {
  _fact[0] = one;
  for (int i = 0; i <= n + 1; i++) _fact[i + 1] = mul_m32(_fact[i], to_m32(i + 1, r2, inv, mod), inv, mod);
  _inv_fact[n + 2] = inv_m32(_fact[n + 2], inv, mod);
  for (int i = n + 2; i > 0; i--) _inv_fact[i - 1] = mul_m32(_inv_fact[i], to_m32(i, r2, inv, mod), inv, mod);
  for (int i = 1; i <= n + 1; i++) _inv_table[i] = mul_m32(_inv_fact[i], _fact[i - 1], inv, mod);
}
m32 *sample_point_shift(m32 *A, int A_len, m32 c) {
  m32 *f = (m32 *)calloc(A_len, sizeof(m32));
  m32 *g = (m32 *)calloc((A_len << 1) - 1, sizeof(m32));
  m32 *ret = (m32 *)calloc(A_len, sizeof(m32));
  #ifdef LOCAL
  if (f == NULL || g == NULL || ret == NULL) exit(EXIT_FAILURE);
  #endif
  for (int i = 0; i < A_len; i++) {
    f[i] = mul_m32(mul_m32(A[i], _inv_fact[i], inv, mod), _inv_fact[A_len - 1 - i], inv, mod);
    if ((A_len - 1 - i) & 1) f[i] = min_m32(f[i], mod);
  }
  for (int i = 0; i < (A_len << 1) - 1; i++) g[i] = inv_m32(mod + c - (A_len - 1) + i, inv, mod);
  m32 *h = convolute_mod1000000007(f, A_len, g, (A_len << 1) - 1);
  m32 coef = one;
  for (int i = 0; i < A_len; i++) coef = mul_m32(coef, c - A_len + 1 + i, inv, mod);
  for (int i = 0; i < A_len; i++) {
    h[i + A_len - 1] = mul_m32(h[i + A_len - 1], coef, inv, mod);
    coef = mul_m32(coef, mul_m32(c + i + 1, g[i], inv, mod), inv, mod);
  }
  for (int i = 0; i < A_len; i++) ret[i] = h[A_len - 1 + i];
  return ret;
}
#pragma endregion sample_point_shift

#pragma region factorial mod
m32 factorial_mod(u64 n) {
  if (n <= 1) return one;
  if (n >= 1000000007ul) return 0;
  const i64 v = 32768;
  m32 iv = inv_m32(to_m32((u32)v, r2, inv, mod), inv, mod);
  m32 *G = (m32 *)calloc(v + 1, sizeof(m32));
  #ifdef LOCAL
  if (G == NULL) exit(EXIT_FAILURE);
  #endif
  G[0] = one;
  G[1] = to_m32((u32)v + 1, r2, inv, mod);
  for (i64 d = 1; d * v < 1000000007; d <<= 1) {
    m32 *G1 = sample_point_shift(G, d + 1, mul_m32(to_m32(d, r2, inv, mod), iv, inv, mod));
    m32 *G2 = sample_point_shift(G, d + 1, mul_m32(to_m32(d * v + v, r2, inv, mod), iv, inv, mod));
    m32 *G3 = sample_point_shift(G, d + 1, mul_m32(to_m32(d * v + d + v, r2, inv, mod), iv, inv, mod));
    for (int i = 0; i <= d; i++) G[i] = mul_m32(G[i], G1[i], inv, mod), G2[i] = mul_m32(G2[i], G3[i], inv, mod);
    for (int i = 0; i < d; i++) G[d + i] = G2[i];
  }
  m32 ret = one;
  i64 i = 0;
  while (i + v <= n) ret = mul_m32(ret, G[i / v], inv, mod), i += v;
  while (i < n) i++, ret = mul_m32(ret, to_m32(i, r2, inv, mod), inv, mod);
  return ret;
}
#pragma endregion factorial mod

void Main(void) {
  pre_fact(1 << 19);
  out_m32(factorial_mod(in_u64()), inv, mod);
  NL();
}

int main(void) {
  Main();
  return 0;
}
yukicoder

結果

テストケース

ソースコード
