Purpose
For operators and arrays satisfying commutativity and associativity and having an identity element, it quickly computes the operation result of a continuous sub-range.
例:
Addition of real numbers forms an Abelian group, so it satisfies the above properties. Specifically:
- Commutativity:
- Associativity:
- Existence of identity element:
Since it satisfies these, it can be loaded onto a Segment Tree to quickly compute the sum of continuous sub-ranges.
Similar data structure: LazySegmentTree
Time Complexity
Initialization: Query:
Usage
Declaration
SegmentTree<class> seg(array_length, lambda_for_binary_operation, identity_element);Initialization
seg.set(i,x);assigns x to the i-th element.
After array construction,
seg.build();builds the segment tree.
Update
After construction,
seg.update(i,x)updates the i-th element to x and rebuilds.
Query
seg.query(a,b)returns the evaluated binary operation over the half-open interval .
Implementation
template <class T>class SegmentTree {public: int n; vector<T> s; const function<T(T, T)> f; const T m; SegmentTree(int _n, const function<T(T, T)> &_f, const T &_m) : f(_f), m(_m) { n = 1; while (n < _n) n <<= 1; s.assign(2 * n, _m); } void set(int k, const T &x) { s[k + n] = x; } void build() { for (int k = n - 1; k > 0; k--) s[k] = f(s[2 * k + 0], s[2 * k + 1]); } void update(int k, const T &x) { k += n; s[k] = x; while (k >>= 1) s[k] = f(s[2 * k + 0], s[2 * k + 1]); } T query(int a, int b) { T L = m, R = m; for (a += n, b += n; a < b; a >>= 1, b >>= 1) { if (a & 1) L = f(L, s[a++]); if (b & 1) R = f(s[--b], R); } return f(L, R); } T operator[](const int &k) const { return s[k + n]; }};Verify
//TODO
Example
To quickly compute the sum of a continuous interval:
SegmentTree<int> seg(n, [](int a, int b){ return a+b; }, 0);
you just need to provide a lambda like this.