7K12 blog

猫でも分かる何か

LIB::WaveletMatrix

ウェーブレット行列
任意区間のK番目の要素を高速に取得するデータ構造
https://ei1333.github.io/library/structure/wavelet/wavelet-matrix.cpp.html

前計算  O(N \log V) クエリ  O(\log V)

  • WaveletMatrix(v): 各要素の高さ v を初期値として構築する.
  • access(k): k 番目の要素を返す.
  • rank(x, r): 区間 [0,r) に含まれる x の個数を返す.
  • kth_smallest(l, r, k): 区間 [l,r) に含まれる要素のうち k 番目(0-indexed) に小さいものを返す.
  • kth_largest(l, r, k): 区間 [l,r)に含まれる要素のうち k 番目 (0-indexed) に大きいものを返す.
  • range_freq(l, r, lower, upper): 区間 [l,r) に含まれる要素のうち [lower,upper) である要素数を返す.
  • prev_value(l, r, upper): 区間 [l,r) に含まれる要素のうち upper の次に小さいものを返す.
  • next_value(l, r, lower): 区間 [l,r) に含まれる要素のうち lower の次に大きいものを返す.
#include <bits/stdc++.h>

namespace LIB
{

	/**
	* @brief 完備辞書 (Succinct Indexable Dictionary)
	**/

	struct SuccinctIndexableDictionary
	{
		using ll = long long;
		ll length = 0;
		ll blocks = 0;
		std::vector< unsigned > bit, sum;

		SuccinctIndexableDictionary() = default;

		SuccinctIndexableDictionary(size_t length) : length(length), blocks((length + 31) >> 5)
		{
			bit.assign(blocks, 0U);
			sum.assign(blocks, 0U);
		}

		bool operator[](ll k) { return (bool((bit[k >> 5] >> (k & 31)) & 1)); }
		ll rank(ll k) { return (ll(sum[k >> 5]) + bitcnt(ll(bit[k >> 5] & ((1U << (k & 31)) - 1)))); }
		ll rank(bool val, ll k) { return (val ? rank(k) : k - rank(k)); }
		void set(ll k) { bit[k >> 5] |= 1U << (k & 31); }

		void build()
		{
			sum[0] = 0U;
			for (ll i = 1; i < blocks; i++) sum[i] = sum[i - 1] + (unsigned)bitcnt(bit[i - 1]);
		}
	private:
		ll bitcnt(ll x) { std::bitset<64> b(x); return b.count(); }
	};

	/*
	 * @brief ウェーブレット行列 (Wavelet Matrix)
	 * @docs docs/wavelet-matrix.md
	 */

	template< typename T, long long MAXLOG > struct WaveletMatrix
	{
		using ll = long long;
		ll length = 0;
		SuccinctIndexableDictionary matrix[MAXLOG] = { 0 };
		ll mid[MAXLOG] = { 0 };

		WaveletMatrix() = default;

		WaveletMatrix(vector< T > v) : length(v.size())
		{
			vector< T > l(length), r(length);
			for (ll level = MAXLOG - 1; level >= 0; level--)
			{
				matrix[level] = SuccinctIndexableDictionary(length + 1);
				ll left = 0, right = 0;
				for (ll i = 0; i < length; i++)
				{
					if (((v[i] >> level) & 1)) matrix[level].set(i), r[right++] = v[i];
					else l[left++] = v[i];
				}
				mid[level] = left;
				matrix[level].build();
				v.swap(l);
				for (ll i = 0; i < right; i++) v[left + i] = r[i];
			}
		}
		pair< ll, ll > succ(bool f, ll l, ll r, ll level) { return { matrix[level].rank(f, l) + mid[level] * f, matrix[level].rank(f, r) + mid[level] * f }; }

		T operator[](const ll& k) { return access(k); }
		// k-th(0-indexed) largest number in v[l,r)
		T kth_largest(ll l, ll r, ll k) { return kth_smallest(l, r, r - l - k - 1); }
		// count i s.t. (l <= i < r) && (lower <= v[i] < upper)
		ll range_freq(ll l, ll r, T lower, T upper) { return range_freq(l, r, upper) - range_freq(l, r, lower); }
		
		// v[k]
		T access(ll k)
		{
			T ret = 0;
			for (ll level = MAXLOG - 1; level >= 0; level--)
			{
				bool f = matrix[level][k];
				if (f) ret |= T(1) << level;
				k = matrix[level].rank(f, k) + mid[level] * f;
			}
			return ret;
		}

		// count i s.t. (0 <= i < r) && v[i] == x
		ll rank(const T& x, ll r)
		{
			ll l = 0;
			for (ll level = MAXLOG - 1; level >= 0; level--) tie(l, r) = succ((x >> level) & 1, l, r, level);
			return r - l;
		}

		// k-th(0-indexed) smallest number in v[l,r)
		T kth_smallest(ll l, ll r, ll k)
		{
			assert(0 <= k && k < r - l);
			T ret = 0;
			for (ll level = MAXLOG - 1; level >= 0; level--)
			{
				ll cnt = matrix[level].rank(false, r) - matrix[level].rank(false, l);
				bool f = cnt <= k;
				if (f) ret |= T(1) << level, k -= cnt;
				tie(l, r) = succ(f, l, r, level);
			}
			return ret;
		}

		// count i s.t. (l <= i < r) && (v[i] < upper)
		ll range_freq(ll l, ll r, T upper)
		{
			ll ret = 0;
			for (ll level = MAXLOG - 1; level >= 0; level--)
			{
				bool f = ((upper >> level) & 1);
				if (f) ret += matrix[level].rank(false, r) - matrix[level].rank(false, l);
				tie(l, r) = succ(f, l, r, level);
			}
			return ret;
		}

		// max v[i] s.t. (l <= i < r) && (v[i] < upper)
		T prev_value(ll l, ll r, T upper)
		{
			ll cnt = range_freq(l, r, upper);
			return cnt == 0 ? T(-1) : kth_smallest(l, r, cnt - 1);
		}

		// min v[i] s.t. (l <= i < r) && (lower <= v[i])
		T next_value(ll l, ll r, T lower)
		{
			ll cnt = range_freq(l, r, lower);
			return cnt == r - l ? T(-1) : kth_smallest(l, r, cnt);
		}
	};

	/**
	* @note
	* 説明 https://ei1333.github.io/library/structure/wavelet/wavelet-matrix.cpp.html
	* 実装例 https://judge.yosupo.jp/submission/72212
	* CompressedWaveletMatrix wm(v);
	* ll a, b, c; cin >> a >> b >> c; vl ans;
	* ans.push_back(wm.kth_smallest(a, b, c));
	**/

	template< typename T = long long, long long MAXLOG = 64ll > struct CompressedWaveletMatrix
	{
		using ll = long long;
		WaveletMatrix< ll, MAXLOG > mat;
		vector< T > ys;

		CompressedWaveletMatrix(const vector< T >& v) : ys(v)
		{
			sort(begin(ys), end(ys));
			ys.erase(unique(begin(ys), end(ys)), end(ys));
			vector< ll > t(v.size());
			for (ll i = 0; i < ll(v.size()); i++) t[i] = get(v[i]);
			mat = WaveletMatrix< ll, MAXLOG >(t);
		}

		T operator[](const ll& k) { return access(k); }
		T access(ll k) { return ys[mat.access(k)]; }
		inline ll get(const T& x) { return lower_bound(begin(ys), end(ys), x) - begin(ys); }
		T kth_smallest(ll l, ll r, ll k) { return ys[mat.kth_smallest(l, r, k)]; }
		T kth_largest(ll l, ll r, ll k) { return ys[mat.kth_largest(l, r, k)]; }
		ll range_freq(ll l, ll r, T upper) { return mat.range_freq(l, r, get(upper)); }
		ll range_freq(ll l, ll r, T lower, T upper) { return mat.range_freq(l, r, get(lower), get(upper)); }

		ll rank(const T& x, ll r)
		{
			auto pos = get(x);
			if (pos == ys.size() || ys[pos] != x) return 0;
			return mat.rank(pos, r);
		}

		T prev_value(ll l, ll r, T upper)
		{
			auto ret = mat.prev_value(l, r, get(upper));
			return ret == -1 ? T(-1) : ys[ret];
		}

		T next_value(ll l, ll r, T lower)
		{
			auto ret = mat.next_value(l, r, get(lower));
			return ret == -1 ? T(-1) : ys[ret];
		}
	};
}

実装例
https://judge.yosupo.jp/submission/72212
https://atcoder.jp/contests/abc234/submissions/28422004