Largest square sub-matrix with equal row, column, and diagonal sum

Given a matrixmat[][] of dimensions N*M, the task is to find the size of the largest square submatrix such that the sum of all rows, columns, diagonals in that submatrix are equal.

Examples:

Input: N = 3, M = 4, mat[][] = [[5, 1, 3, 1], [9, 3, 3, 1], [1, 3, 3, 8]]
Output: 2
Explanation:
The submatrix which satisfies all the given conditions is shown in bold
5 1 3 1
9 3 3 1
1 3 3 8
Therefore, the size of the submatrix is 2.

Input: N = 4, M = 5, mat[][] = [[7, 1, 4, 5, 6], [2, 5, 1, 6, 4], [1, 5, 4, 3, 2], [1, 2, 7, 3, 4]]
Output: 3
Explanation:
The submatrix which satisfies all the given conditions is shown in bold
7 1 4 5 6
2 5 1 6 4
1 5 4 3 2
1 2 7 3 4
Therefore, the size of the submatrix is 3.

Approach:

The given problem can be solved by finding the Prefix Sum of all the rows and the columns and then iterate for all possible sizes of the square submatrix from each cell of the matrix and if there exists any such square matrix that satisfies the given criteria then print that size of a square matrix. Follow the below steps to solve the problem:

• Maintain two prefix sum arrays prefixSumRow[] and prefixSumColumn[] and store the prefix sum of rows and columns of the given matrix respectively.
• Perform the following steps to check if any square matrix starting from the cell (i, j) of size K satisfy the given criteria or not:
  1. Find the sum of elements of primary diagonal of the submatrix mat[i][j] to mat[i + K][j + K] and store it in the variable, say sum.
  2. If the value of the sum is the same as the value mentioned below then return true. Otherwise, return false.
     • The prefix sum of all the rows i.e., the value of prefixSumRow[k][j + K] - prefixSumRow[k][j] for all values of k over then range [i, i + K].
     • The prefix sum of all the columns i.e., the value of prefixSumColumn[i + K][j] - prefixSumColumn[i][k] for all values of k over then range [j, j + K].
     • The prefix sum of anti-diagonal elements.
• Now, iterate for all possible sizes of the square matrix that can be formed over the range [min(N, M), 1] and if there exist any possible satisfies the given criteria using the steps in the above steps, then print that size of a square matrix.

Below is the implementation of the above approach:

3

Time Complexity: O(N*M*min(N, M)²), where N is the number of
Auxiliary Space: O(N*M)