Given a `n x m` matrix, `mat` where each row is sorted in ascending order, produce a `n x 1` vector, `v` where `v[i]` corresponds to some element in `mat[i]` such that the variance of `v` is minimized. The elements of `mat` are unique. \n\nThe only solution I could come up with is a brute force dfs (time complexity `O(n * n^n)`, but that doesn\'t take advantage of each row being sorted nor the matrix having unique elements. Any clue how to solve this problem?

Given a n x m matrix, mat where each row is sorted in ascending order, produce a n x 1 vector, v where v[i] corresponds to some element in mat[i] such that the 