// kod obliczający sumę symboli dwumianow modulo p znaleziony na // https://discuss.codechef.com/questions/84205/how-to-calculate-sum-of-binomial-coefficients-efficiently #include<bits/stdc++.h> #include"message.h" #include"futbol.h" #define _rep(_1, _2, _3, _4, name, ...) name #define rep2(i, n) rep3(i, 0, n) #define rep3(i, a, b) rep4(i, a, b, 1) #define rep4(i, a, b, c) for (int i = int(a); i < int(b); i += int(c)) #define rep(...) _rep(__VA_ARGS__, rep4, rep3, rep2, _)(__VA_ARGS__) using namespace std; using i64 = long long; using u32 = unsigned; using u64 = unsigned long long; using f80 = long double; namespace ntt { using word_t = u32; using dword_t = u64; static const int word_bits = 8 * sizeof(word_t); template <word_t mod, word_t prim_root> class Mod { private: static constexpr word_t mul_inv(word_t n, int e=6, word_t x=1) { return e == 0 ? x : mul_inv(n, e-1, x*(2-x*n)); } public: static constexpr word_t inv = mul_inv(mod); static constexpr word_t r2 = -dword_t(mod) % mod; static constexpr int level = __builtin_ctzll(mod - 1); static_assert(inv * mod == 1, "invalid 1/M modulo 2^@."); Mod() {} Mod(word_t n) : x(init(n)) {}; static word_t modulus() { return mod; } static word_t init(word_t w) { return reduce(dword_t(w) * r2); } static word_t reduce(const dword_t w) { return word_t(w >> word_bits) + mod - word_t((dword_t(word_t(w) * inv) * mod) >> word_bits); } static Mod omega() { return Mod(prim_root).pow((mod - 1) >> level); } Mod& operator += (Mod rhs) { this->x += rhs.x; return *this; } Mod& operator -= (Mod rhs) { this->x += 3 * mod - rhs.x; return *this; } Mod& operator *= (Mod rhs) { this->x = reduce(dword_t(this->x) * rhs.x); return *this; } Mod operator + (Mod rhs) const { return Mod(*this) += rhs; } Mod operator - (Mod rhs) const { return Mod(*this) -= rhs; } Mod operator * (Mod rhs) const { return Mod(*this) *= rhs; } word_t get() const { return reduce(this->x) % mod; } void set(word_t n) const { this->x = n; } Mod pow(word_t exp) const { Mod ret = Mod(1); for (Mod base = *this; exp; exp >>= 1, base *= base) if (exp & 1) ret *= base; return ret; } Mod inverse() const { return pow(mod - 2); } friend ostream& operator << (ostream& os, const Mod& m) { return os << m.get(); } word_t x; }; const int size = 1 << 16; using m32_1 = ntt::Mod<138412033, 5>; using m32_2 = ntt::Mod<155189249, 6>; using m32_3 = ntt::Mod<163577857, 23>; // <= 16579e4 (sub.D = 3) m32_1 f1[size], g1[size]; m32_2 f2[size], g2[size]; m32_3 f3[size], g3[size]; template <typename mod_t> void convolve(mod_t* A, int s1, mod_t* B, int s2, bool cyclic=false) { int s = (cyclic ? max(s1, s2) : s1 + s2 - 1); int size = 1; while (size < s) size <<= 1; mod_t roots[mod_t::level] = { mod_t::omega() }; rep(i, 1, mod_t::level) roots[i] = roots[i - 1] * roots[i - 1]; fill(A + s1, A + size, 0); ntt_dit4(A, size, 1, roots); if (A == B && s1 == s2) { rep(i, size) A[i] *= A[i]; } else { fill(B + s2, B + size, 0); ntt_dit4(B, size, 1, roots); rep(i, size) A[i] *= B[i]; } ntt_dit4(A, size, -1, roots); mod_t inv = mod_t(size).inverse(); rep(i, cyclic ? size : s) A[i] *= inv; } template <typename mod_t> void rev_permute(mod_t* A, int n) { int r = 0, nh = n >> 1; rep(i, 1, n) { for (int h = nh; !((r ^= h) & h); h >>= 1); if (r > i) swap(A[i], A[r]); } } template <typename mod_t> void ntt_dit4(mod_t* A, int n, int sign, mod_t* roots) { rev_permute(A, n); int logn = __builtin_ctz(n); if (logn & 1) rep(i, 0, n, 2) { mod_t a = A[i], b = A[i + 1]; A[i] = a + b; A[i + 1] = a - b; } mod_t imag = roots[mod_t::level - 2]; if (sign < 0) imag = imag.inverse(); mod_t one = mod_t(1); rep(e, 2 + (logn & 1), logn + 1, 2) { const int m = 1 << e; const int m4 = m >> 2; mod_t dw = roots[mod_t::level - e]; if (sign < 0) dw = dw.inverse(); const int block_size = min(n, max(m, (1 << 15) / int(sizeof(A[0])))); rep(k, 0, n, block_size) { mod_t w = one, w2 = one, w3 = one; rep(j, m4) { rep(i, k + j, k + block_size, m) { mod_t a0 = A[i + m4 * 0] * one, a2 = A[i + m4 * 1] * w2; mod_t a1 = A[i + m4 * 2] * w, a3 = A[i + m4 * 3] * w3; mod_t t02 = a0 + a2, t13 = a1 + a3; A[i + m4 * 0] = t02 + t13; A[i + m4 * 2] = t02 - t13; t02 = a0 - a2, t13 = (a1 - a3) * imag; A[i + m4 * 1] = t02 + t13; A[i + m4 * 3] = t02 - t13; } w *= dw; w2 = w * w; w3 = w2 * w; } } } } } using R = int; using R64 = i64; class poly { public: static const int ntt_threshold = 1500; static R add_mod(R a, R b) { return int(a += b - mod) < 0 ? a + mod : a; } static R sub_mod(R a, R b) { return int(a -= b) < 0 ? a + mod : a; } static R64 sub_mul_mod(R64 a, R b, R c) { i64 t = i64(a) - i64(int(b)) * int(c); return t < 0 ? t + lmod : t; } static R mul_mod(R a, R b) { return R64(a) * b % fast_mod; } static R pow_mod(R a, int e) { R ret = 1 % fast_mod; for (; e; e >>= 1, a = mul_mod(a, a)) { if (e & 1) ret = mul_mod(ret, a); } return ret; } static R mod_inv(R a) { R b = mod, s = 1, t = 0; while (b > 0) { swap(s -= t * (a / b), t); swap(a %= b, b); } return int(s) < 0 ? s + mod : s; } inline static void vec_add(R64* res, int s, const R* f, R c) { rep(i, s) res[i] = sub_mul_mod(res[i], mod - c, f[i]); } inline static void vec_sub(R64* res, int s, const R* f, R c) { rep(i, s) res[i] = sub_mul_mod(res[i], c, f[i]); } struct fast_div { fast_div() {} fast_div(u32 n) : m(n) {} friend u32 operator % (u64 n, fast_div d) { return n % d.m; } u32 m; }; public: poly() {} poly(int n) : coefs(n) {} poly(int n, int c) : coefs(n, c % mod) {} poly(const R* ar, int s) : coefs(ar, ar + s) {} poly(const vector<R>& v) : coefs(v) {} poly(const poly& f, int beg, int end=-1) { if (end < 0) end = beg, beg = 0; resize(end - beg); rep(i, beg, end) if (i < f.size()) coefs[i - beg] = f[i]; } static int ilog2(u64 n) { return 63 - __builtin_clzll(n); } int size() const { return coefs.size(); } void resize(int s) { coefs.resize(s); } void push_back(R c) { coefs.push_back(c); } const R* data() const { return coefs.data(); } R* data() { return coefs.data(); } const R& operator [] (int i) const { return coefs[i]; } R& operator [] (int i) { return coefs[i]; } void reverse() { std::reverse(coefs.begin(), coefs.end()); } poly operator - () { poly ret = *this; rep(i, ret.size()) ret[i] = (ret[i] == 0 ? 0 : mod - ret[i]); return ret; } poly& operator += (const poly& rhs) { if (size() < rhs.size()) resize(rhs.size()); rep(i, rhs.size()) coefs[i] = add_mod(coefs[i], rhs[i]); return *this; } poly& operator -= (const poly& rhs) { if (size() < rhs.size()) resize(rhs.size()); rep(i, rhs.size()) coefs[i] = sub_mod(coefs[i], rhs[i]); return *this; } poly& operator *= (const poly& rhs) { return *this = *this * rhs; } poly& rev_add(const poly& rhs) { if (size() < rhs.size()) { int s = size(); resize(rhs.size()); rep(i, s) coefs[size() - 1 - i] = coefs[s - 1 - i]; rep(i, size() - s) coefs[i] = 0; } rep(i, rhs.size()) coefs[size() - 1 - i] = \ add_mod(coefs[size() - 1 - i], rhs.coefs[rhs.size() - 1 - i]); return *this; } poly operator + (const poly& rhs) const { return poly(*this) += rhs; } poly operator - (const poly& rhs) const { return poly(*this) -= rhs; } poly operator * (const poly& rhs) const { return this->mul(rhs); } static void set_mod(R m, int N=2) { mod = m; lmod = R64(m) << 32; N = max(2, N); fast_mod = fast_div(mod); invs.assign(N + 1, 1); facts.assign(N + 1, 1); ifacts.assign(N + 1, 1); invs[1] = 1; rep(i, 2, N + 1) { invs[i] = mul_mod(invs[mod % i], mod - mod / i); facts[i] = mul_mod(facts[i - 1], i); ifacts[i] = mul_mod(ifacts[i - 1], invs[i]); } } private: static poly mul_crt(int beg, int end) { using namespace ntt; auto m1 = m32_1::modulus(); auto m2 = m32_2::modulus(); auto m3 = m32_3::modulus(); auto m12 = u64(m1) * m2; poly ret(end - beg); u32 m12m = m12 % mod; u32 inv1 = m32_2(m1).inverse().get(); u32 inv12 = m32_3(m12 % m3).inverse().get(); rep(i, ret.size()) { u32 r1 = f1[i + beg].get(), r2 = f2[i + beg].get(), r3 = f3[i + beg].get(); u64 r = r1 + u64(r2 + m2 - r1) * inv1 % m2 * m1; ret[i] = (r + u64(r3 + m3 - r % m3) * inv12 % m3 * m12m) % mod; } return ret; } static void mul2(const poly& f, const poly& g, bool cyclic=false) { using namespace ntt; if (&f == &g) { rep(i, f.size()) f1[i] = f[i] % m32_1::modulus(); convolve(f1, f.size(), f1, f.size(), cyclic); rep(i, f.size()) f2[i] = f[i] % m32_2::modulus(); convolve(f2, f.size(), f2, f.size(), cyclic); rep(i, f.size()) f3[i] = f[i] % m32_3::modulus(); convolve(f3, f.size(), f3, f.size(), cyclic); } else { rep(i, f.size()) f1[i] = f[i] % m32_1::modulus(); rep(i, g.size()) g1[i] = g[i] % m32_1::modulus(); convolve(f1, f.size(), g1, g.size(), cyclic); rep(i, f.size()) f2[i] = f[i] % m32_2::modulus(); rep(i, g.size()) g2[i] = g[i] % m32_2::modulus(); convolve(f2, f.size(), g2, g.size(), cyclic); rep(i, f.size()) f3[i] = f[i] % m32_3::modulus(); rep(i, g.size()) g3[i] = g[i] % m32_3::modulus(); convolve(f3, f.size(), g3, g.size(), cyclic); } } public: static void amul(const R* f, int s1, const R* g, int s2, R* res) { int s = s1 + s2 - 1; tmp64.assign(s, 0); rep(i, s2) if (g[i]) vec_add(tmp64.data() + i, s1, f, g[i]); rep(i, s) res[i] = tmp64[i] % fast_mod; } poly mul_basecase(const poly& g) const { const auto& f = *this; int s = size() + g.size() - 1; poly ret(s); amul(f.data(), f.size(), g.data(), g.size(), ret.data()); return ret; } poly mul(const poly& g) const { const auto& f = *this; if (f.size() == 0 || g.size() == 0) return poly(); if (f.size() + g.size() <= ntt_threshold) { return f.mul_basecase(g); } else { mul2(f, g, false); return mul_crt(0, f.size() + g.size() - 1); } } poly middle_product(const poly& g) const { const poly& f = *this; if (f.size() == 0 || g.size() == 0) return poly(); mul2(f, g, true); return mul_crt(f.size(), g.size()); } void print() const { printf("["); if (size()) { printf("%u", coefs[0]); rep(i, 1, size()) printf(", %u", coefs[i]); } puts("]"); } public: vector<R> coefs; static vector<R> tmp32; static vector<R64> tmp64; static vector<R> invs, facts, ifacts; static R mod; static R64 lmod; static fast_div fast_mod; }; R poly::mod; R64 poly::lmod; poly::fast_div poly::fast_mod; vector<R> poly::tmp32; vector<R64> poly::tmp64; vector<R> poly::invs, poly::facts, poly::ifacts; int pow_mod(int b, int e, int mod) { int ret = 1; for (; e; e >>= 1, b = i64(b) * b % mod) { if (e & 1) ret = i64(ret) * b % mod; } return ret; } int binomial_sum_mod_p(int N, int K, int mod) { if (K == 0) return 1 % mod; if (N <= K) return pow_mod(2, N, mod); if (i64(K) * 2 > N) { return (pow_mod(2, N, mod) + i64(mod) - binomial_sum_mod_p(N, N - K - 1, mod)) % mod; } assert(N < mod); const int sqrt_K = sqrt(K); poly::set_mod(mod, sqrt_K); auto mod_invs = [&] (vector<int>& f) { int n = f.size(); vector<int> ret(f); if (n > 0) { rep(i, 1, n) ret[i] = i64(ret[i - 1]) * ret[i] % mod; int inv = poly::mod_inv(ret[n - 1]); for (int i = n - 1; i > 0; --i) { ret[i] = i64(ret[i - 1]) * inv % mod; inv = i64(inv) * f[i] % mod; } ret[0] = inv; } return ret; }; auto conv = [&] (vector<int>& f) -> poly { int n = f.size(); const auto& ifacts = poly::ifacts; auto g = poly(f); rep(i, n) { int d = i64(ifacts[i]) * ifacts[(n - 1) - i] % mod; if ((n - 1 - i) & 1) d = mod - d; g[i] = i64(g[i]) * d % mod; } return g; }; auto shift = [&] (const poly& cf, const poly& f, i64 dx) { if ((dx %= mod) < 0) dx += mod; const int n = f.size(); const int a = i64(dx) * poly::mod_inv(sqrt_K) % mod; auto g = poly(2 * n); rep(i, g.size()) g[i] = (i64(mod) + a + i - n) % mod; rep(i, g.size()) if (g[i] == 0) g[i] = 1; g.coefs = mod_invs(g.coefs); auto ret = cf.middle_product(g); int prod = 1; rep(i, n) prod = i64(prod) * (i64(mod) + a + n - 1 - i) % mod; for (int i = n - 1; i >= 0; --i) { ret[i] = i64(ret[i]) * prod % mod; prod = i64(prod) * g[n + i] % mod * (i64(mod) + a + i - n) % mod; } if (dx % sqrt_K == 0) { int k = n - dx / sqrt_K; rep(i, k) ret[i] = f[n - k + i]; } return ret.coefs; }; using Pair = pair< vector<int>, vector<int> >; function< Pair(int) > rec = [&] (int n) -> Pair { if (n == 1) { return Pair({N, N - sqrt_K}, {1, sqrt_K + 1}); } int nh = n >> 1; auto res = rec(nh); auto& f11 = res.first, &g11 = res.second; auto f = conv(f11), g = conv(g11); auto g12 = shift(g, g11, nh); auto g21 = shift(g, g11, i64(sqrt_K) * nh); auto g22 = shift(g, g11, i64(sqrt_K) * nh + nh); auto f12 = shift(f, f11, N - nh * i64(sqrt_K + 2)); auto f21 = shift(f, f11, i64(sqrt_K) * nh); auto f22 = shift(f, f11, N - i64(2) * nh * (sqrt_K + 1)); rep(i, nh + 1) { g11[i] = (i64(g11[i]) * f12[nh - i] + i64(g12[i]) * f11[i]) % mod; } rep(i, 1, nh + 1) { g11.push_back( (i64(g21[i]) * f22[nh - i] + i64(g22[i]) * f21[i]) % mod ); } f12 = shift(f, f11, nh); f22 = shift(f, f11, i64(sqrt_K) * nh + nh); rep(i, nh + 1) f11[i] = i64(f11[i]) * f12[i] % mod; rep(i, 1, nh + 1) f11.push_back(i64(f21[i]) * f22[i] % mod); if (n & 1) { rep(i, n) { g11[i] = (i64(g11[i]) + f11[i]) * (n + i64(i) * sqrt_K) % mod; } rep(i, n) { f11[i] = i64(f11[i]) * (i64(N) + mod - sqrt_K * i - n + 1) % mod; } vector<int> vals(n); rep(i, n) vals[i] = (i64(sqrt_K) * n + i + 1) % mod; if (i64(sqrt_K + 1) * n < mod) { int prod = 1; rep(i, n) prod = i64(prod) * vals[i] % mod; auto invs = mod_invs(vals); i64 s = 0; rep(i, n) { s += prod; prod = i64(prod) * invs[i] % mod * (i64(N) + mod - i64(sqrt_K) * n - i) % mod; } g11.push_back(s % mod); f11.push_back(prod); } else { g11.push_back(0); f11.push_back(0); } } return {f11, g11}; }; auto res = rec(sqrt_K); auto &f1 = res.first, &g1 = res.second; auto f2 = shift(conv(f1), f1, N - i64(sqrt_K) * (sqrt_K + 1)); reverse(f2.begin(), f2.end()); f2.resize(f2.size() - 1); f2 = mod_invs(f2); i64 ret = 0; rep(i, sqrt_K) { ret = (ret * f1[sqrt_K - 1 - i] + g1[sqrt_K - 1 - i]) % mod; ret = ret * f2[sqrt_K - 1 - i] % mod; } int prod = 1; rep(i, sqrt_K) { prod = i64(prod) * f1[i] % mod * f2[i] % mod; } const int rest = max(0, K - sqrt_K * sqrt_K); ret += prod; vector<int> invs(rest); rep(i, rest) invs[i] = i + 1 + sqrt_K * sqrt_K; invs = mod_invs(invs); rep(i, rest) { prod = i64(prod) * (N - sqrt_K * sqrt_K - i) % mod * invs[i] % mod; ret += prod; } ret %= mod; return ret; } int main() { ios_base::sync_with_stdio(0); cin.tie(0); int n=GetN(), k=GetK(), mod=GetP(), me=MyNodeId(); if(me == 0) { cout << binomial_sum_mod_p(n,k,mod) << "\n"; } }
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 | // kod obliczający sumę symboli dwumianow modulo p znaleziony na // https://discuss.codechef.com/questions/84205/how-to-calculate-sum-of-binomial-coefficients-efficiently #include<bits/stdc++.h> #include"message.h" #include"futbol.h" #define _rep(_1, _2, _3, _4, name, ...) name #define rep2(i, n) rep3(i, 0, n) #define rep3(i, a, b) rep4(i, a, b, 1) #define rep4(i, a, b, c) for (int i = int(a); i < int(b); i += int(c)) #define rep(...) _rep(__VA_ARGS__, rep4, rep3, rep2, _)(__VA_ARGS__) using namespace std; using i64 = long long; using u32 = unsigned; using u64 = unsigned long long; using f80 = long double; namespace ntt { using word_t = u32; using dword_t = u64; static const int word_bits = 8 * sizeof(word_t); template <word_t mod, word_t prim_root> class Mod { private: static constexpr word_t mul_inv(word_t n, int e=6, word_t x=1) { return e == 0 ? x : mul_inv(n, e-1, x*(2-x*n)); } public: static constexpr word_t inv = mul_inv(mod); static constexpr word_t r2 = -dword_t(mod) % mod; static constexpr int level = __builtin_ctzll(mod - 1); static_assert(inv * mod == 1, "invalid 1/M modulo 2^@."); Mod() {} Mod(word_t n) : x(init(n)) {}; static word_t modulus() { return mod; } static word_t init(word_t w) { return reduce(dword_t(w) * r2); } static word_t reduce(const dword_t w) { return word_t(w >> word_bits) + mod - word_t((dword_t(word_t(w) * inv) * mod) >> word_bits); } static Mod omega() { return Mod(prim_root).pow((mod - 1) >> level); } Mod& operator += (Mod rhs) { this->x += rhs.x; return *this; } Mod& operator -= (Mod rhs) { this->x += 3 * mod - rhs.x; return *this; } Mod& operator *= (Mod rhs) { this->x = reduce(dword_t(this->x) * rhs.x); return *this; } Mod operator + (Mod rhs) const { return Mod(*this) += rhs; } Mod operator - (Mod rhs) const { return Mod(*this) -= rhs; } Mod operator * (Mod rhs) const { return Mod(*this) *= rhs; } word_t get() const { return reduce(this->x) % mod; } void set(word_t n) const { this->x = n; } Mod pow(word_t exp) const { Mod ret = Mod(1); for (Mod base = *this; exp; exp >>= 1, base *= base) if (exp & 1) ret *= base; return ret; } Mod inverse() const { return pow(mod - 2); } friend ostream& operator << (ostream& os, const Mod& m) { return os << m.get(); } word_t x; }; const int size = 1 << 16; using m32_1 = ntt::Mod<138412033, 5>; using m32_2 = ntt::Mod<155189249, 6>; using m32_3 = ntt::Mod<163577857, 23>; // <= 16579e4 (sub.D = 3) m32_1 f1[size], g1[size]; m32_2 f2[size], g2[size]; m32_3 f3[size], g3[size]; template <typename mod_t> void convolve(mod_t* A, int s1, mod_t* B, int s2, bool cyclic=false) { int s = (cyclic ? max(s1, s2) : s1 + s2 - 1); int size = 1; while (size < s) size <<= 1; mod_t roots[mod_t::level] = { mod_t::omega() }; rep(i, 1, mod_t::level) roots[i] = roots[i - 1] * roots[i - 1]; fill(A + s1, A + size, 0); ntt_dit4(A, size, 1, roots); if (A == B && s1 == s2) { rep(i, size) A[i] *= A[i]; } else { fill(B + s2, B + size, 0); ntt_dit4(B, size, 1, roots); rep(i, size) A[i] *= B[i]; } ntt_dit4(A, size, -1, roots); mod_t inv = mod_t(size).inverse(); rep(i, cyclic ? size : s) A[i] *= inv; } template <typename mod_t> void rev_permute(mod_t* A, int n) { int r = 0, nh = n >> 1; rep(i, 1, n) { for (int h = nh; !((r ^= h) & h); h >>= 1); if (r > i) swap(A[i], A[r]); } } template <typename mod_t> void ntt_dit4(mod_t* A, int n, int sign, mod_t* roots) { rev_permute(A, n); int logn = __builtin_ctz(n); if (logn & 1) rep(i, 0, n, 2) { mod_t a = A[i], b = A[i + 1]; A[i] = a + b; A[i + 1] = a - b; } mod_t imag = roots[mod_t::level - 2]; if (sign < 0) imag = imag.inverse(); mod_t one = mod_t(1); rep(e, 2 + (logn & 1), logn + 1, 2) { const int m = 1 << e; const int m4 = m >> 2; mod_t dw = roots[mod_t::level - e]; if (sign < 0) dw = dw.inverse(); const int block_size = min(n, max(m, (1 << 15) / int(sizeof(A[0])))); rep(k, 0, n, block_size) { mod_t w = one, w2 = one, w3 = one; rep(j, m4) { rep(i, k + j, k + block_size, m) { mod_t a0 = A[i + m4 * 0] * one, a2 = A[i + m4 * 1] * w2; mod_t a1 = A[i + m4 * 2] * w, a3 = A[i + m4 * 3] * w3; mod_t t02 = a0 + a2, t13 = a1 + a3; A[i + m4 * 0] = t02 + t13; A[i + m4 * 2] = t02 - t13; t02 = a0 - a2, t13 = (a1 - a3) * imag; A[i + m4 * 1] = t02 + t13; A[i + m4 * 3] = t02 - t13; } w *= dw; w2 = w * w; w3 = w2 * w; } } } } } using R = int; using R64 = i64; class poly { public: static const int ntt_threshold = 1500; static R add_mod(R a, R b) { return int(a += b - mod) < 0 ? a + mod : a; } static R sub_mod(R a, R b) { return int(a -= b) < 0 ? a + mod : a; } static R64 sub_mul_mod(R64 a, R b, R c) { i64 t = i64(a) - i64(int(b)) * int(c); return t < 0 ? t + lmod : t; } static R mul_mod(R a, R b) { return R64(a) * b % fast_mod; } static R pow_mod(R a, int e) { R ret = 1 % fast_mod; for (; e; e >>= 1, a = mul_mod(a, a)) { if (e & 1) ret = mul_mod(ret, a); } return ret; } static R mod_inv(R a) { R b = mod, s = 1, t = 0; while (b > 0) { swap(s -= t * (a / b), t); swap(a %= b, b); } return int(s) < 0 ? s + mod : s; } inline static void vec_add(R64* res, int s, const R* f, R c) { rep(i, s) res[i] = sub_mul_mod(res[i], mod - c, f[i]); } inline static void vec_sub(R64* res, int s, const R* f, R c) { rep(i, s) res[i] = sub_mul_mod(res[i], c, f[i]); } struct fast_div { fast_div() {} fast_div(u32 n) : m(n) {} friend u32 operator % (u64 n, fast_div d) { return n % d.m; } u32 m; }; public: poly() {} poly(int n) : coefs(n) {} poly(int n, int c) : coefs(n, c % mod) {} poly(const R* ar, int s) : coefs(ar, ar + s) {} poly(const vector<R>& v) : coefs(v) {} poly(const poly& f, int beg, int end=-1) { if (end < 0) end = beg, beg = 0; resize(end - beg); rep(i, beg, end) if (i < f.size()) coefs[i - beg] = f[i]; } static int ilog2(u64 n) { return 63 - __builtin_clzll(n); } int size() const { return coefs.size(); } void resize(int s) { coefs.resize(s); } void push_back(R c) { coefs.push_back(c); } const R* data() const { return coefs.data(); } R* data() { return coefs.data(); } const R& operator [] (int i) const { return coefs[i]; } R& operator [] (int i) { return coefs[i]; } void reverse() { std::reverse(coefs.begin(), coefs.end()); } poly operator - () { poly ret = *this; rep(i, ret.size()) ret[i] = (ret[i] == 0 ? 0 : mod - ret[i]); return ret; } poly& operator += (const poly& rhs) { if (size() < rhs.size()) resize(rhs.size()); rep(i, rhs.size()) coefs[i] = add_mod(coefs[i], rhs[i]); return *this; } poly& operator -= (const poly& rhs) { if (size() < rhs.size()) resize(rhs.size()); rep(i, rhs.size()) coefs[i] = sub_mod(coefs[i], rhs[i]); return *this; } poly& operator *= (const poly& rhs) { return *this = *this * rhs; } poly& rev_add(const poly& rhs) { if (size() < rhs.size()) { int s = size(); resize(rhs.size()); rep(i, s) coefs[size() - 1 - i] = coefs[s - 1 - i]; rep(i, size() - s) coefs[i] = 0; } rep(i, rhs.size()) coefs[size() - 1 - i] = \ add_mod(coefs[size() - 1 - i], rhs.coefs[rhs.size() - 1 - i]); return *this; } poly operator + (const poly& rhs) const { return poly(*this) += rhs; } poly operator - (const poly& rhs) const { return poly(*this) -= rhs; } poly operator * (const poly& rhs) const { return this->mul(rhs); } static void set_mod(R m, int N=2) { mod = m; lmod = R64(m) << 32; N = max(2, N); fast_mod = fast_div(mod); invs.assign(N + 1, 1); facts.assign(N + 1, 1); ifacts.assign(N + 1, 1); invs[1] = 1; rep(i, 2, N + 1) { invs[i] = mul_mod(invs[mod % i], mod - mod / i); facts[i] = mul_mod(facts[i - 1], i); ifacts[i] = mul_mod(ifacts[i - 1], invs[i]); } } private: static poly mul_crt(int beg, int end) { using namespace ntt; auto m1 = m32_1::modulus(); auto m2 = m32_2::modulus(); auto m3 = m32_3::modulus(); auto m12 = u64(m1) * m2; poly ret(end - beg); u32 m12m = m12 % mod; u32 inv1 = m32_2(m1).inverse().get(); u32 inv12 = m32_3(m12 % m3).inverse().get(); rep(i, ret.size()) { u32 r1 = f1[i + beg].get(), r2 = f2[i + beg].get(), r3 = f3[i + beg].get(); u64 r = r1 + u64(r2 + m2 - r1) * inv1 % m2 * m1; ret[i] = (r + u64(r3 + m3 - r % m3) * inv12 % m3 * m12m) % mod; } return ret; } static void mul2(const poly& f, const poly& g, bool cyclic=false) { using namespace ntt; if (&f == &g) { rep(i, f.size()) f1[i] = f[i] % m32_1::modulus(); convolve(f1, f.size(), f1, f.size(), cyclic); rep(i, f.size()) f2[i] = f[i] % m32_2::modulus(); convolve(f2, f.size(), f2, f.size(), cyclic); rep(i, f.size()) f3[i] = f[i] % m32_3::modulus(); convolve(f3, f.size(), f3, f.size(), cyclic); } else { rep(i, f.size()) f1[i] = f[i] % m32_1::modulus(); rep(i, g.size()) g1[i] = g[i] % m32_1::modulus(); convolve(f1, f.size(), g1, g.size(), cyclic); rep(i, f.size()) f2[i] = f[i] % m32_2::modulus(); rep(i, g.size()) g2[i] = g[i] % m32_2::modulus(); convolve(f2, f.size(), g2, g.size(), cyclic); rep(i, f.size()) f3[i] = f[i] % m32_3::modulus(); rep(i, g.size()) g3[i] = g[i] % m32_3::modulus(); convolve(f3, f.size(), g3, g.size(), cyclic); } } public: static void amul(const R* f, int s1, const R* g, int s2, R* res) { int s = s1 + s2 - 1; tmp64.assign(s, 0); rep(i, s2) if (g[i]) vec_add(tmp64.data() + i, s1, f, g[i]); rep(i, s) res[i] = tmp64[i] % fast_mod; } poly mul_basecase(const poly& g) const { const auto& f = *this; int s = size() + g.size() - 1; poly ret(s); amul(f.data(), f.size(), g.data(), g.size(), ret.data()); return ret; } poly mul(const poly& g) const { const auto& f = *this; if (f.size() == 0 || g.size() == 0) return poly(); if (f.size() + g.size() <= ntt_threshold) { return f.mul_basecase(g); } else { mul2(f, g, false); return mul_crt(0, f.size() + g.size() - 1); } } poly middle_product(const poly& g) const { const poly& f = *this; if (f.size() == 0 || g.size() == 0) return poly(); mul2(f, g, true); return mul_crt(f.size(), g.size()); } void print() const { printf("["); if (size()) { printf("%u", coefs[0]); rep(i, 1, size()) printf(", %u", coefs[i]); } puts("]"); } public: vector<R> coefs; static vector<R> tmp32; static vector<R64> tmp64; static vector<R> invs, facts, ifacts; static R mod; static R64 lmod; static fast_div fast_mod; }; R poly::mod; R64 poly::lmod; poly::fast_div poly::fast_mod; vector<R> poly::tmp32; vector<R64> poly::tmp64; vector<R> poly::invs, poly::facts, poly::ifacts; int pow_mod(int b, int e, int mod) { int ret = 1; for (; e; e >>= 1, b = i64(b) * b % mod) { if (e & 1) ret = i64(ret) * b % mod; } return ret; } int binomial_sum_mod_p(int N, int K, int mod) { if (K == 0) return 1 % mod; if (N <= K) return pow_mod(2, N, mod); if (i64(K) * 2 > N) { return (pow_mod(2, N, mod) + i64(mod) - binomial_sum_mod_p(N, N - K - 1, mod)) % mod; } assert(N < mod); const int sqrt_K = sqrt(K); poly::set_mod(mod, sqrt_K); auto mod_invs = [&] (vector<int>& f) { int n = f.size(); vector<int> ret(f); if (n > 0) { rep(i, 1, n) ret[i] = i64(ret[i - 1]) * ret[i] % mod; int inv = poly::mod_inv(ret[n - 1]); for (int i = n - 1; i > 0; --i) { ret[i] = i64(ret[i - 1]) * inv % mod; inv = i64(inv) * f[i] % mod; } ret[0] = inv; } return ret; }; auto conv = [&] (vector<int>& f) -> poly { int n = f.size(); const auto& ifacts = poly::ifacts; auto g = poly(f); rep(i, n) { int d = i64(ifacts[i]) * ifacts[(n - 1) - i] % mod; if ((n - 1 - i) & 1) d = mod - d; g[i] = i64(g[i]) * d % mod; } return g; }; auto shift = [&] (const poly& cf, const poly& f, i64 dx) { if ((dx %= mod) < 0) dx += mod; const int n = f.size(); const int a = i64(dx) * poly::mod_inv(sqrt_K) % mod; auto g = poly(2 * n); rep(i, g.size()) g[i] = (i64(mod) + a + i - n) % mod; rep(i, g.size()) if (g[i] == 0) g[i] = 1; g.coefs = mod_invs(g.coefs); auto ret = cf.middle_product(g); int prod = 1; rep(i, n) prod = i64(prod) * (i64(mod) + a + n - 1 - i) % mod; for (int i = n - 1; i >= 0; --i) { ret[i] = i64(ret[i]) * prod % mod; prod = i64(prod) * g[n + i] % mod * (i64(mod) + a + i - n) % mod; } if (dx % sqrt_K == 0) { int k = n - dx / sqrt_K; rep(i, k) ret[i] = f[n - k + i]; } return ret.coefs; }; using Pair = pair< vector<int>, vector<int> >; function< Pair(int) > rec = [&] (int n) -> Pair { if (n == 1) { return Pair({N, N - sqrt_K}, {1, sqrt_K + 1}); } int nh = n >> 1; auto res = rec(nh); auto& f11 = res.first, &g11 = res.second; auto f = conv(f11), g = conv(g11); auto g12 = shift(g, g11, nh); auto g21 = shift(g, g11, i64(sqrt_K) * nh); auto g22 = shift(g, g11, i64(sqrt_K) * nh + nh); auto f12 = shift(f, f11, N - nh * i64(sqrt_K + 2)); auto f21 = shift(f, f11, i64(sqrt_K) * nh); auto f22 = shift(f, f11, N - i64(2) * nh * (sqrt_K + 1)); rep(i, nh + 1) { g11[i] = (i64(g11[i]) * f12[nh - i] + i64(g12[i]) * f11[i]) % mod; } rep(i, 1, nh + 1) { g11.push_back( (i64(g21[i]) * f22[nh - i] + i64(g22[i]) * f21[i]) % mod ); } f12 = shift(f, f11, nh); f22 = shift(f, f11, i64(sqrt_K) * nh + nh); rep(i, nh + 1) f11[i] = i64(f11[i]) * f12[i] % mod; rep(i, 1, nh + 1) f11.push_back(i64(f21[i]) * f22[i] % mod); if (n & 1) { rep(i, n) { g11[i] = (i64(g11[i]) + f11[i]) * (n + i64(i) * sqrt_K) % mod; } rep(i, n) { f11[i] = i64(f11[i]) * (i64(N) + mod - sqrt_K * i - n + 1) % mod; } vector<int> vals(n); rep(i, n) vals[i] = (i64(sqrt_K) * n + i + 1) % mod; if (i64(sqrt_K + 1) * n < mod) { int prod = 1; rep(i, n) prod = i64(prod) * vals[i] % mod; auto invs = mod_invs(vals); i64 s = 0; rep(i, n) { s += prod; prod = i64(prod) * invs[i] % mod * (i64(N) + mod - i64(sqrt_K) * n - i) % mod; } g11.push_back(s % mod); f11.push_back(prod); } else { g11.push_back(0); f11.push_back(0); } } return {f11, g11}; }; auto res = rec(sqrt_K); auto &f1 = res.first, &g1 = res.second; auto f2 = shift(conv(f1), f1, N - i64(sqrt_K) * (sqrt_K + 1)); reverse(f2.begin(), f2.end()); f2.resize(f2.size() - 1); f2 = mod_invs(f2); i64 ret = 0; rep(i, sqrt_K) { ret = (ret * f1[sqrt_K - 1 - i] + g1[sqrt_K - 1 - i]) % mod; ret = ret * f2[sqrt_K - 1 - i] % mod; } int prod = 1; rep(i, sqrt_K) { prod = i64(prod) * f1[i] % mod * f2[i] % mod; } const int rest = max(0, K - sqrt_K * sqrt_K); ret += prod; vector<int> invs(rest); rep(i, rest) invs[i] = i + 1 + sqrt_K * sqrt_K; invs = mod_invs(invs); rep(i, rest) { prod = i64(prod) * (N - sqrt_K * sqrt_K - i) % mod * invs[i] % mod; ret += prod; } ret %= mod; return ret; } int main() { ios_base::sync_with_stdio(0); cin.tie(0); int n=GetN(), k=GetK(), mod=GetP(), me=MyNodeId(); if(me == 0) { cout << binomial_sum_mod_p(n,k,mod) << "\n"; } } |