#####   ######  #####    ###    #   #  ###  #   #  ######
##  ##  ##      ##  ##  ## ##   #   #   #   #   #  ##
#####   ####    #####   #   #   #   #   #   #   #  ####
##  #   ##      ##      ## ##    # #    #    # #   ##
##   #  ######  ##       ###      #    ###    #    ######

$ curl repovive.com/roadmaps/data-structures/fenwick-trees/2d-range-sum-query

░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░█████████████████████████████████████████████████████████████████████████████

#####   ######  #####    ###    #   #  ###  #   #  ######
##  ##  ##      ##  ##  ## ##   #   #   #   #   #  ##
#####   ####    #####   #   #   #   #   #   #   #  ####
##  #   ##      ##      ## ##    # #    #    # #   ##
##   #  ######  ##       ###      #    ###    #    ######

$ curl repovive.com/roadmaps/data-structures/fenwick-trees/2d-range-sum-query

Repovive

2D Range Sum Query - Fenwick Trees | Data Structures | Repovive

Repovive.

Data Structures19 sections · 729 units49h total

Open in Course

2D Range Sum Query

Rectangle queries

To query sum over rectangle $(r_1, c_1)$ to $(r_2, c_2)$ , use inclusion-exclusion:

$\text{sum} = Q(r_2, c_2) - Q(r_1-1, c_2) - Q(r_2, c_1-1) + Q(r_1-1, c_1-1)$

where $Q(x, y)$ is the prefix sum to $(x, y)$ .

def rangeSum(r1, c1, r2, c2):
    return (query2D(r2, c2)
          - query2D(r1-1, c2)
          - query2D(r2, c1-1)
          + query2D(r1-1, c1-1))

You're applying the same inclusion-exclusion principle used for 2D prefix sums, but now with $O(\log n \log m)$ queries and updates instead of $O(1)$ queries with $O(nm)$ updates.