[Medium] 221. Maximal Square
Given an m x n binary matrix filled with '0's and '1's, find the largest square containing only '1's and return its area.
Examples
Example 1:
Input:
1 0 1 0 0
1 0 1 1 1
1 1 1 1 1
1 0 0 1 0
Output: 4 (a 2x2 square)
Example 2:
Input:
0 1
1 0
Output: 1 (a 1x1 square)
Example 3:
Input:
0
Output: 0
Constraints
m == matrix.lengthn == matrix[i].length1 <= m, n <= 300matrix[i][j]is'0'or'1'
Thinking Process
The DP Definition
Define dp[i][j] = side length of the largest square whose bottom-right corner is at (i, j).
The Transition
A square at (i, j) can only be as large as the smallest of its three neighbors plus one:
dp[i][j] = 1 + min(dp[i-1][j], dp[i][j-1], dp[i-1][j-1])
Why? A square of side k at (i, j) requires:
- A square of side
k-1ending at(i-1, j)(top) - A square of side
k-1ending at(i, j-1)(left) - A square of side
k-1ending at(i-1, j-1)(diagonal)
If any of these is smaller, it becomes the bottleneck.
┌──────────┐
│ diag top│
│ left cur│
└──────────┘
dp[i-1][j-1] dp[i-1][j]
dp[i][j-1] dp[i][j] = 1 + min(top, left, diag)
Complete Walk-through
Matrix: DP table:
1 0 1 0 0 1 0 1 0 0
1 0 1 1 1 1 0 1 1 1
1 1 1 1 1 1 1 1 2 2
1 0 0 1 0 1 0 0 1 0
Key cells:
dp[2][3]:min(top=1, left=1, diag=1) + 1 = 2– a 2x2 square formsdp[2][4]:min(top=1, left=2, diag=1) + 1 = 2– another 2x2 squaredp[3][3]:min(top=2, left=0, diag=1) + 1 = 1– left is 0, so only 1x1
Max value = 2, so answer = 2^2 = 4.
Common Approaches
Typical techniques for this pattern:
| Approach | Time | Space | Notes |
|---|---|---|---|
| 1D DP (this problem) | O(n) | O(n) or O(1) | Linear recurrence |
| 2D DP | O(nm) | O(nm) or O(n) | Grid or two-sequence problems |
| State machine DP | O(n) | O(1) | Buy/sell, hold/not-hold states |
| Memoization (top-down) | Same as DP | O(n) | Recursive + cache |
Solution
from typing import List
class Solution:
def maximalSquare(self, matrix: List[List[str]]) -> int:
if not matrix:
return 0
rows, cols = len(matrix), len(matrix[0])
dp = [[0] * cols for _ in range(rows)]
max_side = 0
for i in range(rows):
for j in range(cols):
if matrix[i][j] == "1":
if i == 0 or j == 0:
dp[i][j] = 1
else:
dp[i][j] = 1 + min(
dp[i - 1][j],
dp[i][j - 1],
dp[i - 1][j - 1],
)
max_side = max(max_side, dp[i][j])
return max_side * max_side
Solution Explanation
Approach: 1D DP (this problem)
Key idea: ### The DP Definition
How the code works:
- A square of side
k-1ending at(i-1, j)(top) - A square of side
k-1ending at(i, j-1)(left) - A square of side
k-1ending at(i-1, j-1)(diagonal) dp[2][3]:min(top=1, left=1, diag=1) + 1 = 2– a 2x2 square formsdp[2][4]:min(top=1, left=2, diag=1) + 1 = 2– another 2x2 squaredp[3][3]:min(top=2, left=0, diag=1) + 1 = 1– left is 0, so only 1x1
Walkthrough — input 1 0 1 0 0, expected output 4 (a 2x2 square):
- Initialize variables from the problem setup.
- Apply the main loop / recursion until the condition is met.
- Confirm the result matches the expected output.
Common Mistakes
- Returning
maxSideinstead ofmaxSide * maxSide: The problem asks for area, not side length - Treating
matrix[i][j]as int: The matrix containscharvalues ('0'/'1'), not integers - Forgetting to reset
dp[j] = 0in 1D version: Whenmatrix[i][j] == '0', the cell must be explicitly zeroed
Key Takeaways
- Classic 2D DP pattern: define state at each cell, derive from neighbors
- The
minof three neighbors is the core insight – a square is only as large as its weakest constraint - Space optimization from 2D to 1D is a standard technique: save the diagonal before overwriting
- Answer is text{maxSide}^2 (area, not side length)
Related Problems
- 85. Maximal Rectangle – harder generalization using histogram approach
- 1277. Count Square Submatrices with All Ones – same DP, sum all
dp[i][j]values - 62. Unique Paths – similar 2D DP grid pattern
- 64. Minimum Path Sum – 2D DP with neighbor transitions
References
- LC 221: Maximal Square on LeetCode
- LeetCode Discuss — LC 221: Maximal Square
- LeetCode Editorial (may require premium)