2D Range Queries

Authors: Benjamin Qi, Andi Qu

Extending Range Queries to 2D (and beyond).


Quite rare, I've only needed this once.


Focus Problem – try your best to solve this problem before continuing!



Essentially, we just nest the loops that one would find in a 1D BIT to get N-dimensional BITs. We can then use PIE to query subrectangles.


#include <bits/stdc++.h>
using namespace std;
int bit[1001][1001];
int n;
void update(int x, int y, int val) {
for (; x <= n; x += (x & (-x))) {
for (int i = y; i <= n; i += (i & (-i))) { bit[x][i] += val; }

Alternative Implementation

Using the multidimensional implementation mentioned here.

template <class T, int... Ns> struct BIT {
T val = 0;
void upd(T v) { val += v; }
T query() { return val; }
template <class T, int N, int... Ns> struct BIT<T, N, Ns...> {
BIT<T, Ns...> bit[N + 1];
template <typename... Args> void upd(int pos, Args... args) {
for (; pos <= N; pos += (pos & -pos)) bit[pos].upd(args...);


StatusSourceProblem NameDifficultyTags
Back From SummerNormal
Show Tags2DRQ, BIT
Show Tags3DRQ, BIT

Optional: Range Update and Range Query in Higher Dimensions

Lazy propagation on segment trees does not extend to higher dimensions. However, you can extend the 1D BIT solution to solve range increment range sum in higher dimensions as well! See this paper for details.

2D Offline Sum Queries

See my implementations.

Focus Problem – try your best to solve this problem before continuing!

The intended complexity is O(Nlog2N)\mathcal{O}(N\log^2 N) with a good constant factor. This requires updating points and querying rectangle sums NN times for points with coordinates in the range [1,N][1,N]. However, the 2D BITs mentioned above use O(N2)\mathcal{O}(N^2) memory, which is too much.

Solution - Soriya's Programming Project

Since we know all of the updates and queries beforehand, we can reduce the memory usage while maintaining a decent constant factor.

Idea 1 - Use an unordered map instead of a 2D array.

Bad idea ... This gives O(Nlog2N)\mathcal{O}(N\log^2N) memory and time and the constant factors for both are terrible.

Idea 2 - Compress the points to be updated so that you only need O(NlogN)\mathcal{O}(N\log N) memory.

This doesn't require knowing the queries beforehand.

It's a bit difficult to pass the above problem within the time limit. Make sure to use fast input (and not endl)!

Idea 3 - Use divide & conquer with a 1D BIT

The fastest way.


StatusSourceProblem NameDifficultyTags
Show Tags2DRQ, BIT
Show Tags2DRQ, BIT
Show Tags2DRQ, BIT

2D Segment Tree

A segment tree of (maybe sparse) segment trees.

Pro Tip

This is not the same as Quadtree. If the coordinates go up to CC, then 2D segment tree queries run in O(log2C)\mathcal{O}(\log^2C) time each but some queries make Quadtree take Θ(C)\Theta(C) time!


Brief description





More code

Note - Memory Usage

Naively, inserting NN elements into a sparse segment tree requires O(NlogC)\mathcal{O}(N\log C) memory, giving a bound of O(Nlog2C)\mathcal{O}(N\log^2C) on 2D segment tree memory. This is usually too much for N=C=105N=C=10^5 and 256 MB (although it sufficed for "Mowing the Field" due to the 512MB memory limit). Possible ways to get around this:

  • Use arrays of fixed size rather than pointers.
  • Reduce the memory usage of sparse segment tree to O(N)\mathcal{O}(N) while maintaining the same O(NlogC)\mathcal{O}(N\log C) insertion time (see the solution for IOI Game below for details).
  • Use BBSTs instead of sparse segment trees. O(N)\mathcal{O}(N) memory, O(NlogN)\mathcal{O}(N\log N) insertion time.


Can also try the USACO problems from above.

StatusSourceProblem NameDifficultyTags
Show Tags2DRQ, Lazy SegTree
Show Tags2DRQ, Sparse SegTree, Treap
JOIVery Hard
Show Tags2DRQ, SegTree

Module Progress:

Join the USACO Forum!

Stuck on a problem, or don't understand a module? Join the USACO Forum and get help from other competitive programmers!