Matrix Questions at Yahoo: What to Expect

Matrix questions appear in roughly 11% of Yahoo's technical interview problems. For a company handling vast datasets, search infrastructure, and advertising systems, matrices are a natural structure for representing grids, images, adjacency in graphs, and tabular data. Success here demonstrates your ability to navigate 2D data efficiently—a core skill for backend and data-oriented roles.

What to Expect — Types of Problems

Yahoo's matrix problems typically focus on practical applications over abstract math. Expect these categories:

1. Traversal & Search: Zigzag traversal, searching in a row-wise and column-wise sorted matrix (common at Yahoo), or finding a specific element.
2. Dynamic Programming in Grids: Problems involving finding paths (often with obstacles) or calculating minimum/maximum sum paths across a matrix.
3. Modification In-Place: Rotating images (a classic), or setting matrix rows and columns to zero based on cell values. These test your understanding of space complexity.
4. Island/Region Problems: Counting connected components (e.g., number of islands) using DFS or BFS on a grid, which models real-world segmentation tasks.

The difficulty is usually medium. Interviewers will look for a correct, optimal solution and clean code.

How to Prepare — Study Tips with One Code Example

Focus on pattern recognition. Don't memorize problems—learn the underlying techniques: systematic traversal (nested loops, layer-by-layer), DFS/BFS for connectivity, and DP for optimization on grids. Always clarify edge cases: empty matrix, 1x1 matrix, or single row/column.

A key pattern is searching in a sorted matrix. This is efficient and frequently asked. The optimal approach starts from the top-right (or bottom-left) corner, eliminating an entire row or column with each comparison.

def search_sorted_matrix(matrix, target):
    if not matrix:
        return False
    rows, cols = len(matrix), len(matrix[0])
    # Start from top-right corner
    r, c = 0, cols - 1
    while r < rows and c >= 0:
        if matrix[r][c] == target:
            return True
        elif matrix[r][c] > target:
            c -= 1  # Eliminate column
        else:
            r += 1  # Eliminate row
    return False

Recommended Practice Order

Build competence sequentially:

1. Start with fundamental traversals: spiral order, diagonal traversal.
2. Move to classic modification problems: set matrix zeroes, rotate image.
3. Practice pathfinding and DP: minimum path sum, unique paths.
4. Tackle DFS/BFS applications: number of islands, flood fill.
5. Finally, solve search and optimization in sorted matrices.

This order builds from basic control flow to advanced graph and optimization concepts, ensuring you have the tools for any variant Yahoo presents.

Practice Matrix at Yahoo