Find a Peak Element II

Description:​

A peak element in a 2D grid is an element that is strictly greater than all of its adjacent neighbors to the left, right, top, and bottom.

Given a 0-indexed m x n matrix mat where no two adjacent cells are equal, find any peak element mat[i][j] and return the length 2 array [i,j].

You may assume that the entire matrix is surrounded by an outer perimeter with the value -1 in every cell.

You must write an algorithm that runs in $O(M \log N)$ or $O(N \log M)$ time.

Example 1:

Input: mat = [[1,4],[3,2]] Output: [0,1] Explanation: Both 3 and 4 are peak elements so [1,0] and [0,1] are both acceptable answers.

Example 2:

Input: mat = [[10,20,15],[21,30,14],[7,16,32]] Output: [1,1] Explanation: Both 30 and 32 are peak elements so [1,1] and [2,2] are both acceptable answers.

Video Explanation:​

Approaches:​

1. Optimal Approach (Binary Search)​

Since we are required to solve this in $O(M \log N)$ or $O(N \log M)$ time, a linear scan of the whole matrix $O(M \times N)$ is not acceptable. We must use Binary Search.

We can perform a binary search on the columns (from $0$ to $N-1$). At each step, we look at the middle column mid.

1. Find Global Maximum in Column: Iterate through all rows in the mid column to find the row index (maxRow) containing the maximum value.
2. Evaluate Neighbors: Since mat[maxRow][mid] is the largest element in its column, it is guaranteed to be strictly greater than its top and bottom neighbors. Now, we only need to compare it with its left and right neighbors.
   • If it is greater than both its left and right neighbors, we found a peak! Return [maxRow, mid].
   • If its left neighbor is strictly greater, it means there must be a peak somewhere in the left half of the matrix. We eliminate the right half by setting high = mid - 1.
   • If its right neighbor is strictly greater, it means there must be a peak somewhere in the right half of the matrix. We eliminate the left half by setting low = mid + 1.

Complexity​

• Time Complexity: $O(M \log N)$ where $M$ is the number of rows and $N$ is the number of columns. We perform a binary search on the columns ($\log N$) and at each step, we scan through $M$ rows to find the maximum element.
• Space Complexity: $O(1)$ as we only use a few pointers and integer variables.

Solutions:​

C++

class Solution {
public:
    int findMaxRow(vector<vector<int>>& mat, int col) {
        int maxVal = -1;
        int maxRow = -1;
        for (int i = 0; i < mat.size(); i++) {
            if (mat[i][col] > maxVal) {
                maxVal = mat[i][col];
                maxRow = i;
            }
        }
        return maxRow;
    }

    vector<int> findPeakGrid(vector<vector<int>>& mat) {
        int n = mat.size();
        int m = mat[0].size();
        int low = 0, high = m - 1;

        while (low <= high) {
            int mid = low + (high - low) / 2;
            int maxRow = findMaxRow(mat, mid);

            int leftVal = mid - 1 >= 0 ? mat[maxRow][mid - 1] : -1;
            int rightVal = mid + 1 < m ? mat[maxRow][mid + 1] : -1;

            if (mat[maxRow][mid] > leftVal && mat[maxRow][mid] > rightVal) {
                return {maxRow, mid};
            } else if (mat[maxRow][mid] < leftVal) {
                high = mid - 1;
            } else {
                low = mid + 1;
            }
        }
        return {-1, -1};
    }
};

Java

class Solution {
    public int findMaxRow(int[][] mat, int col) {
        int maxVal = -1;
        int maxRow = -1;
        for (int i = 0; i < mat.length; i++) {
            if (mat[i][col] > maxVal) {
                maxVal = mat[i][col];
                maxRow = i;
            }
        }
        return maxRow;
    }

    public int[] findPeakGrid(int[][] mat) {
        int n = mat.length;
        int m = mat[0].length;
        int low = 0, high = m - 1;

        while (low <= high) {
            int mid = low + (high - low) / 2;
            int maxRow = findMaxRow(mat, mid);

            int leftVal = mid - 1 >= 0 ? mat[maxRow][mid - 1] : -1;
            int rightVal = mid + 1 < m ? mat[maxRow][mid + 1] : -1;

            if (mat[maxRow][mid] > leftVal && mat[maxRow][mid] > rightVal) {
                return new int[]{maxRow, mid};
            } else if (mat[maxRow][mid] < leftVal) {
                high = mid - 1;
            } else {
                low = mid + 1;
            }
        }
        return new int[]{-1, -1};
    }
}

Python

class Solution:
    def findPeakGrid(self, mat: list[list[int]]) -> list[int]:
        n, m = len(mat), len(mat[0])
        low, high = 0, m - 1
        
        while low <= high:
            mid = low + (high - low) // 2
            max_row = 0
            for i in range(n):
                if mat[i][mid] > mat[max_row][mid]:
                    max_row = i
                    
            left_val = mat[max_row][mid - 1] if mid - 1 >= 0 else -1
            right_val = mat[max_row][mid + 1] if mid + 1 < m else -1
            
            if mat[max_row][mid] > left_val and mat[max_row][mid] > right_val:
                return [max_row, mid]
            elif mat[max_row][mid] < left_val:
                high = mid - 1
            else:
                low = mid + 1
                
        return [-1, -1]

JavaScript

/**
 * @param {number[][]} mat
 * @return {number[]}
 */
var findPeakGrid = function(mat) {
    let n = mat.length;
    let m = mat[0].length;
    let low = 0, high = m - 1;
    
    while (low <= high) {
        let mid = Math.floor(low + (high - low) / 2);
        
        let maxRow = 0;
        for (let i = 0; i < n; i++) {
            if (mat[i][mid] > mat[maxRow][mid]) {
                maxRow = i;
            }
        }
        
        let leftVal = mid - 1 >= 0 ? mat[maxRow][mid - 1] : -1;
        let rightVal = mid + 1 < m ? mat[maxRow][mid + 1] : -1;
        
        if (mat[maxRow][mid] > leftVal && mat[maxRow][mid] > rightVal) {
            return [maxRow, mid];
        } else if (mat[maxRow][mid] < leftVal) {
            high = mid - 1;
        } else {
            low = mid + 1;
        }
    }
    
    return [-1, -1];
};