結果
問題 | No.1783 Remix Sum |
ユーザー | Pachicobue |
提出日時 | 2021-12-12 21:38:53 |
言語 | C++17 (gcc 12.3.0 + boost 1.83.0) |
結果 |
WA
|
実行時間 | - |
コード長 | 25,624 bytes |
コンパイル時間 | 2,501 ms |
コンパイル使用メモリ | 227,108 KB |
実行使用メモリ | 58,500 KB |
最終ジャッジ日時 | 2024-07-21 09:51:26 |
合計ジャッジ時間 | 48,244 ms |
ジャッジサーバーID (参考情報) |
judge3 / judge1 |
(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 | 1 ms
5,248 KB |
testcase_01 | AC | 2 ms
5,248 KB |
testcase_02 | AC | 2 ms
5,376 KB |
testcase_03 | WA | - |
testcase_04 | AC | 248 ms
5,516 KB |
testcase_05 | AC | 168 ms
5,512 KB |
testcase_06 | WA | - |
testcase_07 | AC | 706 ms
12,208 KB |
testcase_08 | WA | - |
testcase_09 | AC | 419 ms
8,620 KB |
testcase_10 | AC | 168 ms
5,516 KB |
testcase_11 | AC | 89 ms
5,376 KB |
testcase_12 | AC | 941 ms
13,492 KB |
testcase_13 | AC | 244 ms
10,940 KB |
testcase_14 | AC | 279 ms
7,720 KB |
testcase_15 | WA | - |
testcase_16 | AC | 707 ms
12,212 KB |
testcase_17 | AC | 945 ms
12,852 KB |
testcase_18 | AC | 87 ms
5,376 KB |
testcase_19 | WA | - |
testcase_20 | WA | - |
testcase_21 | WA | - |
testcase_22 | WA | - |
testcase_23 | AC | 547 ms
8,416 KB |
testcase_24 | WA | - |
testcase_25 | WA | - |
testcase_26 | WA | - |
testcase_27 | WA | - |
testcase_28 | AC | 226 ms
8,212 KB |
testcase_29 | AC | 417 ms
8,292 KB |
testcase_30 | WA | - |
testcase_31 | AC | 558 ms
8,600 KB |
testcase_32 | WA | - |
testcase_33 | AC | 414 ms
8,420 KB |
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 | WA | - |
testcase_43 | WA | - |
testcase_44 | AC | 6,610 ms
37,208 KB |
testcase_45 | WA | - |
testcase_46 | TLE | - |
testcase_47 | -- | - |
testcase_48 | -- | - |
testcase_49 | -- | - |
testcase_50 | -- | - |
testcase_51 | -- | - |
testcase_52 | -- | - |
testcase_53 | -- | - |
testcase_54 | -- | - |
testcase_55 | -- | - |
testcase_56 | -- | - |
testcase_57 | -- | - |
testcase_58 | -- | - |
testcase_59 | -- | - |
testcase_60 | -- | - |
testcase_61 | -- | - |
testcase_62 | -- | - |
testcase_63 | -- | - |
testcase_64 | -- | - |
testcase_65 | -- | - |
testcase_66 | -- | - |
testcase_67 | -- | - |
testcase_68 | -- | - |
testcase_69 | -- | - |
testcase_70 | -- | - |
testcase_71 | -- | - |
testcase_72 | -- | - |
testcase_73 | -- | - |
testcase_74 | -- | - |
testcase_75 | -- | - |
testcase_76 | -- | - |
testcase_77 | -- | - |
testcase_78 | -- | - |
testcase_79 | -- | - |
ソースコード
#include <bits/stdc++.h> /** * 上位は1の10乗根でアダマール変換をして、下位はNTTする? * 多変数FPSでexp.logを経てM乗を計算? * https://nyaannyaan.github.io/library/ntt/multivariate-multiplication.hpp */ #pragma region Header using i32 = int; using u32 = unsigned int; using i64 = long long; using u64 = unsigned long long; using i128 = __int128_t; using u128 = __uint128_t; using f64 = double; using f80 = long double; using f128 = __float128; constexpr i32 operator"" _i32(u64 v) { return v; } constexpr i32 operator"" _u32(u64 v) { return v; } constexpr i64 operator"" _i64(u64 v) { return v; } constexpr u64 operator"" _u64(u64 v) { return v; } constexpr f64 operator"" _f64(f80 v) { return v; } constexpr f80 operator"" _f80(f80 v) { return v; } using Istream = std::istream; using Ostream = std::ostream; using Str = std::string; template<typename T> using Lt = std::less<T>; template<typename T> using Gt = std::greater<T>; template<typename T> using IList = std::initializer_list<T>; template<int n> using BSet = std::bitset<n>; template<typename T1, typename T2> using Pair = std::pair<T1, T2>; template<typename... Ts> using Tup = std::tuple<Ts...>; template<typename T, int N> using Arr = std::array<T, N>; template<typename... Ts> using Deq = std::deque<Ts...>; template<typename... Ts> using Set = std::set<Ts...>; template<typename... Ts> using MSet = std::multiset<Ts...>; template<typename... Ts> using USet = std::unordered_set<Ts...>; template<typename... Ts> using UMSet = std::unordered_multiset<Ts...>; template<typename... Ts> using Map = std::map<Ts...>; template<typename... Ts> using MMap = std::multimap<Ts...>; template<typename... Ts> using UMap = std::unordered_map<Ts...>; template<typename... Ts> using UMMap = std::unordered_multimap<Ts...>; template<typename... Ts> using Vec = std::vector<Ts...>; template<typename... Ts> using Stack = std::stack<Ts...>; template<typename... Ts> using Queue = std::queue<Ts...>; template<typename T> using MaxHeap = std::priority_queue<T>; template<typename T> using MinHeap = std::priority_queue<T, Vec<T>, Gt<T>>; using NSec = std::chrono::nanoseconds; using USec = std::chrono::microseconds; using MSec = std::chrono::milliseconds; using Sec = std::chrono::seconds; template<typename T> constexpr T LIMMIN = std::numeric_limits<T>::min(); template<typename T> constexpr T LIMMAX = std::numeric_limits<T>::max(); template<typename T> constexpr T INF = (LIMMAX<T> - 1) / 2; template<typename T> constexpr T PI = T{3.141592653589793238462643383279502884}; template<typename T = u64> constexpr T TEN(const int n) { return n == 0 ? T{1} : TEN<T>(n - 1) * T{10}; } Ostream& operator<<(Ostream& os, i128 v) { bool minus = false; if (v < 0) { minus = true, v = -v; } Str ans; if (v == 0) { ans = "0"; } while (v) { ans.push_back('0' + v % 10), v /= 10; } std::reverse(ans.begin(), ans.end()); return os << (minus ? "-" : "") << ans; } Ostream& operator<<(Ostream& os, u128 v) { Str ans; if (v == 0) { ans = "0"; } while (v) { ans.push_back('0' + v % 10), v /= 10; } std::reverse(ans.begin(), ans.end()); return os << ans; } template<typename T> bool chmin(T& a, const T& b) { if (a > b) { a = b; return true; } else { return false; } } template<typename T> bool chmax(T& a, const T& b) { if (a < b) { a = b; return true; } else { return false; } } template<typename T> constexpr T floorDiv(T x, T y) { if (y < T{}) { x = -x, y = -y; } return x >= T{} ? x / y : (x - y + 1) / y; } template<typename T> constexpr T ceilDiv(T x, T y) { if (y < T{}) { x = -x, y = -y; } return x >= T{} ? (x + y - 1) / y : x / y; } template<typename T, typename I> constexpr T modPower(T v, I n, T mod) { T ans = 1 % mod; for (; n > 0; n >>= 1, (v *= v) %= mod) { if (n % 2 == 1) { (ans *= v) %= mod; } } return ans; } template<typename T, typename I> constexpr T power(T v, I n) { T ans = 1; for (; n > 0; n >>= 1, v *= v) { if (n % 2 == 1) { ans *= v; } } return ans; } template<typename T, typename I> constexpr T power(T v, I n, const T& e) { T ans = e; for (; n > 0; n >>= 1, v *= v) { if (n % 2 == 1) { ans *= v; } } return ans; } template<typename T> Vec<T> operator+=(Vec<T>& vs1, const Vec<T>& vs2) { vs1.insert(vs1.end(), vs2.begin(), vs2.end()); return vs1; } template<typename T> Vec<T> operator+(const Vec<T>& vs1, const Vec<T>& vs2) { auto vs = vs1; vs += vs2; return vs; } template<typename Vs, typename V> void fillAll(Vs& arr, const V& v) { if constexpr (std::is_convertible<V, Vs>::value) { arr = v; } else { for (auto& subarr : arr) { fillAll(subarr, v); } } } template<typename Vs> void sortAll(Vs& vs) { std::sort(std::begin(vs), std::end(vs)); } template<typename Vs, typename C> void sortAll(Vs& vs, C comp) { std::sort(std::begin(vs), std::end(vs), comp); } template<typename Vs> void reverseAll(Vs& vs) { std::reverse(std::begin(vs), std::end(vs)); } template<typename V, typename Vs> V sumAll(const Vs& vs) { if constexpr (std::is_convertible<Vs, V>::value) { return static_cast<V>(vs); } else { V ans = 0; for (const auto& v : vs) { ans += sumAll<V>(v); } return ans; } } template<typename Vs> int minInd(const Vs& vs) { return std::min_element(std::begin(vs), std::end(vs)) - std::begin(vs); } template<typename Vs> int maxInd(const Vs& vs) { return std::max_element(std::begin(vs), std::end(vs)) - std::begin(vs); } template<typename Vs, typename V> int lbInd(const Vs& vs, const V& v) { return std::lower_bound(std::begin(vs), std::end(vs), v) - std::begin(vs); } template<typename Vs, typename V> int ubInd(const Vs& vs, const V& v) { return std::upper_bound(std::begin(vs), std::end(vs), v) - std::begin(vs); } template<typename T, typename F> Vec<T> genVec(int n, F gen) { Vec<T> ans; std::generate_n(std::back_insert_iterator(ans), n, gen); return ans; } Vec<int> iotaVec(int n, int offset = 0) { Vec<int> ans(n); std::iota(ans.begin(), ans.end(), offset); return ans; } constexpr int popcount(const u64 v) { return v ? __builtin_popcountll(v) : 0; } constexpr int log2p1(const u64 v) { return v ? 64 - __builtin_clzll(v) : 0; } constexpr int lsbp1(const u64 v) { return __builtin_ffsll(v); } constexpr int clog(const u64 v) { return v ? log2p1(v - 1) : 0; } constexpr u64 ceil2(const u64 v) { const int l = clog(v); return (l == 64) ? 0_u64 : (1_u64 << l); } constexpr u64 floor2(const u64 v) { return v ? (1_u64 << (log2p1(v) - 1)) : 0_u64; } constexpr bool ispow2(const u64 v) { return (v > 0) and ((v & (v - 1)) == 0); } constexpr bool btest(const u64 mask, const int ind) { return (mask >> ind) & 1_u64; } template<typename F> struct Fix : F { Fix(F&& f) : F{std::forward<F>(f)} {} template<typename... Args> auto operator()(Args&&... args) const { return F::operator()(*this, std::forward<Args>(args)...); } }; class irange { private: struct itr { itr(i64 start = 0, i64 step = 1) : m_cnt{start}, m_step{step} {} bool operator!=(const itr& it) const { return m_cnt != it.m_cnt; } int operator*() { return m_cnt; } itr& operator++() { m_cnt += m_step; return *this; } i64 m_cnt, m_step; }; i64 m_start, m_end, m_step; public: irange(i64 start, i64 end, i64 step = 1) { assert(step != 0); const i64 d = std::abs(step); const i64 l = (step > 0 ? start : end); const i64 r = (step > 0 ? end : start); int n = (r - l) / d + ((r - l) % d ? 1 : 0); if (l >= r) { n = 0; } m_start = start; m_end = start + step * n; m_step = step; } itr begin() const { return itr{m_start, m_step}; } itr end() const { return itr{m_end, m_step}; } }; irange rep(int end) { return irange(0, end, 1); } irange per(int rend) { return irange(rend - 1, -1, -1); } #pragma COMMENT("[REFS] Xoshiro: https://prng.di.unimi.it") namespace xoshiro_impl { u64 x; u64 next() { uint64_t z = (x += 0x9e3779b97f4a7c15); z = (z ^ (z >> 30)) * 0xbf58476d1ce4e5b9; z = (z ^ (z >> 27)) * 0x94d049bb133111eb; return z ^ (z >> 31); } } // namespace xoshiro_impl class Xoshiro32 { public: using result_type = u32; using T = result_type; Xoshiro32(T seed = 0) { xoshiro_impl::x = seed; s[0] = xoshiro_impl::next(); s[1] = xoshiro_impl::next(); s[2] = xoshiro_impl::next(); s[3] = xoshiro_impl::next(); } static constexpr T min() { return LIMMIN<T>; } static constexpr T max() { return LIMMAX<T>; } T operator()() { return next(); } private: static constexpr T rotl(const T x, int k) { return (x << k) | (x >> (32 - k)); } T next() { const T ans = rotl(s[1] * 5, 7) * 9; const T t = s[1] << 9; s[2] ^= s[0]; s[3] ^= s[1]; s[1] ^= s[2]; s[0] ^= s[3]; s[2] ^= t; s[3] = rotl(s[3], 11); return ans; } T s[4]; }; class Xoshiro64 { public: using result_type = u64; using T = result_type; Xoshiro64(T seed = 0) { xoshiro_impl::x = seed; s[0] = xoshiro_impl::next(); s[1] = xoshiro_impl::next(); s[2] = xoshiro_impl::next(); s[3] = xoshiro_impl::next(); } static constexpr T min() { return LIMMIN<T>; } static constexpr T max() { return LIMMAX<T>; } T operator()() { return next(); } private: static constexpr T rotl(const T x, int k) { return (x << k) | (x >> (64 - k)); } T next() { const T ans = rotl(s[1] * 5, 7) * 9; const T t = s[1] << 17; s[2] ^= s[0]; s[3] ^= s[1]; s[1] ^= s[2]; s[0] ^= s[3]; s[2] ^= t; s[3] = rotl(s[3], 45); return ans; } T s[4]; }; template<typename Rng> class RNG { public: using result_type = typename Rng::result_type; using T = result_type; static constexpr T min() { return Rng::min(); } static constexpr T max() { return Rng::max(); } RNG() : RNG(std::random_device{}()) {} RNG(T seed) : m_rng(seed) {} T operator()() { return m_rng(); } template<typename T> T val(T min, T max) { return std::uniform_int_distribution<T>(min, max)(m_rng); } template<typename T> Pair<T, T> pair(T min, T max) { return std::minmax({val<T>(min, max), val<T>(min, max)}); } template<typename T> Vec<T> vec(int n, T min, T max) { return genVec<T>(n, [&]() { return val<T>(min, max); }); } template<typename T> Vec<Vec<T>> vvec(int n, int m, T min, T max) { return genVec<Vec<T>>(n, [&]() { return vec(m, min, max); }); } private: Rng m_rng; }; RNG<std::mt19937> rng; RNG<std::mt19937_64> rng64; RNG<Xoshiro32> rng_xo; RNG<Xoshiro64> rng_xo64; class Scanner { public: Scanner(Istream& is = std::cin) : m_is{is} { m_is.tie(nullptr)->sync_with_stdio(false); } template<typename T> T val() { T v; return m_is >> v, v; } template<typename T> T val(T offset) { return val<T>() - offset; } template<typename T> Vec<T> vec(int n) { return genVec<T>(n, [&]() { return val<T>(); }); } template<typename T> Vec<T> vec(int n, T offset) { return genVec<T>(n, [&]() { return val<T>(offset); }); } template<typename T> Vec<Vec<T>> vvec(int n, int m) { return genVec<Vec<T>>(n, [&]() { return vec<T>(m); }); } template<typename T> Vec<Vec<T>> vvec(int n, int m, const T offset) { return genVec<Vec<T>>(n, [&]() { return vec<T>(m, offset); }); } template<typename... Args> auto tup() { return Tup<Args...>{val<Args>()...}; } template<typename... Args> auto tup(const Args&... offsets) { return Tup<Args...>{val<Args>(offsets)...}; } private: Istream& m_is; }; Scanner in; class Printer { public: Printer(Ostream& os = std::cout) : m_os{os} { m_os << std::fixed << std::setprecision(15); } template<typename... Args> int operator()(const Args&... args) { dump(args...); return 0; } template<typename... Args> int ln(const Args&... args) { dump(args...), m_os << '\n'; return 0; } template<typename... Args> int el(const Args&... args) { dump(args...), m_os << std::endl; return 0; } private: template<typename T> void dump(const T& v) { m_os << v; } template<typename T> void dump(const Vec<T>& vs) { for (const int i : rep(vs.size())) { m_os << (i ? " " : ""), dump(vs[i]); } } template<typename T> void dump(const Vec<Vec<T>>& vss) { for (const int i : rep(vss.size())) { m_os << (i ? "\n" : ""), dump(vss[i]); } } template<typename T, typename... Ts> int dump(const T& v, const Ts&... args) { dump(v), m_os << ' ', dump(args...); return 0; } Ostream& m_os; }; Printer out; template<u32 mod_, u32 root_, u32 max2p_> class modint { template<typename U = u32&> static U modRef() { static u32 s_mod = 0; return s_mod; } template<typename U = u32&> static U rootRef() { static u32 s_root = 0; return s_root; } template<typename U = u32&> static U max2pRef() { static u32 s_max2p = 0; return s_max2p; } public: template<typename U = const u32> static constexpr std::enable_if_t<mod_ != 0, U> mod() { return mod_; } template<typename U = const u32> static std::enable_if_t<mod_ == 0, U> mod() { return modRef(); } template<typename U = const u32> static constexpr std::enable_if_t<mod_ != 0, U> root() { return root_; } template<typename U = const u32> static std::enable_if_t<mod_ == 0, U> root() { return rootRef(); } template<typename U = const u32> static constexpr std::enable_if_t<mod_ != 0, U> max2p() { return max2p_; } template<typename U = const u32> static std::enable_if_t<mod_ == 0, U> max2p() { return max2pRef(); } template<typename U = u32> static void setMod(std::enable_if_t<mod_ == 0, U> m) { modRef() = m; } template<typename U = u32> static void setRoot(std::enable_if_t<mod_ == 0, U> r) { rootRef() = r; } template<typename U = u32> static void setMax2p(std::enable_if_t<mod_ == 0, U> m) { max2pRef() = m; } constexpr modint() : m_val{0} {} constexpr modint(i64 v) : m_val{normll(v)} {} constexpr void setRaw(u32 v) { m_val = v; } constexpr modint operator-() const { return modint{0} - (*this); } constexpr modint& operator+=(const modint& m) { m_val = norm(m_val + m.val()); return *this; } constexpr modint& operator-=(const modint& m) { m_val = norm(m_val + mod() - m.val()); return *this; } constexpr modint& operator*=(const modint& m) { m_val = normll((i64)m_val * (i64)m.val() % (i64)mod()); return *this; } constexpr modint& operator/=(const modint& m) { return *this *= m.inv(); } constexpr modint operator+(const modint& m) const { auto v = *this; return v += m; } constexpr modint operator-(const modint& m) const { auto v = *this; return v -= m; } constexpr modint operator*(const modint& m) const { auto v = *this; return v *= m; } constexpr modint operator/(const modint& m) const { auto v = *this; return v /= m; } constexpr bool operator==(const modint& m) const { return m_val == m.val(); } constexpr bool operator!=(const modint& m) const { return not(*this == m); } friend Istream& operator>>(Istream& is, modint& m) { i64 v; return is >> v, m = v, is; } friend Ostream& operator<<(Ostream& os, const modint& m) { return os << m.val(); } constexpr u32 val() const { return m_val; } template<typename I> constexpr modint pow(I n) const { return power(*this, n); } constexpr modint inv() const { return pow(mod() - 2); } static modint sinv(u32 n) { static Vec<modint> is{1, 1}; for (u32 i = (u32)is.size(); i <= n; i++) { is.push_back(-is[mod() % i] * (mod() / i)); } return is[n]; } static modint fact(u32 n) { static Vec<modint> fs{1, 1}; for (u32 i = (u32)fs.size(); i <= n; i++) { fs.push_back(fs.back() * i); } return fs[n]; } static modint ifact(u32 n) { static Vec<modint> ifs{1, 1}; for (u32 i = (u32)ifs.size(); i <= n; i++) { ifs.push_back(ifs.back() * sinv(i)); } return ifs[n]; } static modint comb(int n, int k) { return k > n or k < 0 ? modint{0} : fact(n) * ifact(n - k) * ifact(k); } private: static constexpr u32 norm(u32 x) { return x < mod() ? x : x - mod(); } static constexpr u32 normll(i64 x) { return norm(u32(x % (i64)mod() + (i64)mod())); } u32 m_val; }; using modint_1000000007 = modint<1000000007, 5, 1>; using modint_998244353 = modint<998244353, 3, 23>; template<int id> using modint_dynamic = modint<0, 0, id>; #pragma endregion constexpr int D = 10; constexpr int Ds[] = {1, 10, 100, 1000, 10000, 100000}; constexpr int ls[] = {0, 4, 7, 10, 14, 17}; constexpr int Ls[] = {1, 16, 128, 1024, 16384, 131072}; constexpr u32 MOD = (1_u32 << 20) * 115 + 1; constexpr u32 ROOT = 6; constexpr u32 MAX2P = 20; using mint = modint<MOD, ROOT, MAX2P>; using fps = Vec<mint>; const mint omega_10 = mint(ROOT).pow((MOD - 1) / 10); const mint omega_10s[] = {1, omega_10, omega_10.pow(2), omega_10.pow(3), omega_10.pow(4), omega_10.pow(5), omega_10.pow(6), omega_10.pow(7), omega_10.pow(8), omega_10.pow(9)}; const mint i_omega_10 = mint(ROOT).pow((MOD - 1) / 10).inv(); const mint i_omega_10s[] = {1, i_omega_10, i_omega_10.pow(2), i_omega_10.pow(3), i_omega_10.pow(4), i_omega_10.pow(5), i_omega_10.pow(6), i_omega_10.pow(7), i_omega_10.pow(8), i_omega_10.pow(9)}; i64 M; int K; // 桁数 int T; // (mod x^D)で計算する桁数、残りは(mod x^D-1)で計算 int L1; // ceil2(D^T) int l1; // log2(L1) int L2; // D^(K-T) int l2; // log10(L2)=K-T int L; // L1*L2 やり取りする長さ /** * A:長さLをNTTする(上位は保ったまま) */ void ntt(Vec<mint>& as, int lg, bool rev) { assert((int)as.size() == L); const int N1 = (1 << lg); assert(N1 == L1); static Vec<mint> rs, irs; if (rs.empty()) { const mint r = mint(mint::root()), ir = r.inv(); rs.resize(mint::max2p() + 1), irs.resize(mint::max2p() + 1); rs.back() = -r.pow((mint::mod() - 1) >> mint::max2p()), irs.back() = -ir.pow((mint::mod() - 1) >> mint::max2p()); for (u32 i : irange(mint::max2p(), 0, -1)) { rs[i - 1] = -(rs[i] * rs[i]); irs[i - 1] = -(irs[i] * irs[i]); } } const auto drange = (rev ? irange(0, lg, 1) : irange(lg - 1, -1, -1)); for (const int d : drange) { const int width = 1 << d; mint e = 1; for (int i = 0, j = 1; i < N1; i += width * 2, j++) { for (int l = i, r = i + width; l < i + width; l++, r++) { for (int x2 : rep(L2)) { const int L = x2 * L1 + l; const int R = x2 * L1 + r; if (rev) { const mint x = as[L], y = as[R]; as[L] = x + y, as[R] = (x - y) * e; } else { const mint x = as[L], y = as[R] * e; as[L] = x + y, as[R] = x - y; } } } e *= (rev ? irs : rs)[lsbp1(j) + 1]; } } if (rev) { const mint iN = mint{N1}.inv(); for (auto& a : as) { a *= iN; } } } /* xにおけるbaseに対応する桁 */ int btest_10(int x, int base) { return (x / base) % D; } /** * A:長さLをFHT_10する(下位は保ったまま) */ void fht_10(Vec<mint>& as, int lg_10, bool rev) { assert((int)as.size() == L); const int N2 = Ds[lg_10]; assert(N2 == L2); for (int delta = 1; delta < N2; delta *= D) { for (int j : rep(N2)) { if (btest_10(j, delta) == 0) { for (int x1 : rep(L1)) { Vec<mint> dps(D); for (int k : rep(D)) { dps[k] = as[(j + delta * k) * L1 + x1]; as[(j + delta * k) * L1 + x1] = 0; } for (int k : rep(D)) { for (int l : rep(D)) { as[(j + delta * k) * L1 + x1] += (rev ? i_omega_10s[(k * l) % D] : omega_10s[(k * l) % D]) * dps[l]; } } } } } } if (rev) { const mint iN = mint(N2).inv(); for (auto& a : as) { a *= iN; } } } /** * A:長さ10^(K-T)をFHT_10する */ void trans(Vec<mint>& as, bool rev) { assert((int)as.size() == L); if (rev) { fht_10(as, l2, rev); ntt(as, l1, rev); } else { ntt(as, l1, rev); fht_10(as, l2, rev); } } /** * f(x1,x2,...,xK)*g(x1,x2,...,xK) * modは(x1^D), ..., (xT^D) : (x_{T+1}^D-1), ..., (xK^D-1) */ fps multi_mul(const fps& f, const fps& g) { assert((int)f.size() == L); assert((int)g.size() == L); if (T == 0) { // 全部FHT auto F = f, G = g; trans(F, false); trans(G, false); for (int x : rep(L)) { F[x] *= G[x]; } trans(F, true); return F; } else { Vec<int> chi(L1, 0); for (int x1 : rep(L1)) { for (int k : irange(1, T)) { chi[x1] += x1 / Ds[k]; } chi[x1] %= T; // mod (t^T-1) } Vec<fps> F(T, fps(L, 0)), G(T, fps(L, 0)); for (int x : rep(L)) { const int x1 = x % L1; F[chi[x1]][x] += f[x]; G[chi[x1]][x] += g[x]; } for (int k : rep(T)) { trans(F[k], false); trans(G[k], false); } // F(t),G(t)の各点積 for (int x : rep(L)) { fps H_x(T, 0); // H_x(t) = F_x(t)*G_x(t) mod (t^T-1) for (int ki : rep(T)) { for (int kj : rep(T)) { H_x[(ki + kj) % T] += F[ki][x] * G[kj][x]; } } for (int k : rep(T)) { F[k][x] = H_x[k]; } } for (int k : rep(T)) { trans(F[k], true); } fps h(L, 0); for (int x : rep(L)) { const int x1 = x % L1; h[x] += F[chi[x1]][x]; } return h; } } /** * f(x1,x2,...,xK)^M * modは(x1^D), ..., (xT^D) : (x_{T+1}^D-1), ..., (xK^D-1) */ fps multi_pow(fps f, i64 M) { if (M == 1) { return f; } else if (M % 2 == 0) { return multi_pow(multi_mul(f, f), M / 2); } else { return multi_mul(f, multi_pow(f, M - 1)); } } int main() { int N; std::tie(N, K, M, T) = in.tup<int, int, i64, int>(); L1 = Ls[T]; l1 = ls[T]; assert(L1 == (1 << l1)); L2 = Ds[K - T]; l2 = K - T; L = L1 * L2; void(0); const auto as = in.vec<int>(N); fps f(L, 0); for (int i : rep(N)) { const int x1 = as[i] % Ds[T]; const int x2 = as[i] / Ds[T]; f[x2 * L1 + x1] += 1; } const auto g = multi_pow(f, M); void(0); Vec<mint> ans(Ds[K], 0); for (int x : rep(L)) { const int x1 = x % L1; const int x2 = x / L1; if (x1 >= Ds[T] or x2 >= Ds[K - T]) { continue; } ans[x2 * Ds[T] + x1] += g[x]; } for (auto an : ans) { out.ln(an); } return 0; }