In a 2D grid of 0
s and 1
s, we change at most one 0
to a 1
.html
After, what is the size of the largest island? (An island is a 4-directionally connected group of 1
s).java
Example 1:git
Input: [[1, 0], [0, 1]] Output: 3 Explanation: Change one 0 to 1 and connect two 1s, then we get an island with area = 3.
Example 2:github
Input: [[1, 1], [1, 0]] Output: 4 Explanation: Change the 0 to 1 and make the island bigger, only one island with area = 4.
Example 3:code
Input: [[1, 1], [1, 1]] Output: 4 Explanation: Can't change any 0 to 1, only one island with area = 4.
Notes:htm
1 <= grid.length = grid[0].length <= 50
.0 <= grid[i][j] <= 1
.
這道題在只有0和1的矩陣中用相連的1來表示小島,如今說是有一個把0變爲1的機會,這樣本來不相鄰的島就有可能變的相鄰了,從而組成一個更大的島,讓求出可能組成的最大的島嶼的面積,也就是相連的1的個數。在 LeetCode 中關於島嶼的題其實也作過許多,好比 Number of Distinct Islands II,Max Area of Island,Number of Distinct Islands,Number of Islands II,和 Number of Islands。其實大多題目的本質都是用 DFS 或者 BFS 去遍歷全部相連的1,固然這道題也不例外。博主最開始的想法是首先用 DFS 來找出每一個島嶼,而後把同一個島嶼的位置座標都放到同一個 HashSet 中,這樣就有了不少 HashSet,而後遍歷全部的0的位置,對每一個0位置,遍歷其周圍4個鄰居,而後看鄰居位置有沒有屬於島嶼的,有的話就把該島嶼的 HashSet 編號記錄下來,遍歷完4個鄰居後,在把全部的相連的島嶼中的1個數加起來(由於 HashSet 能夠直接求出集合中數字的總個數),每次更新結果 res 便可。這種方法是能夠經過 OJ 的,速度還比下面展現的兩種方法要快,就是代碼比較長,沒有下面方法的簡潔,這裏就不貼了。下面的這兩種方法其實都是從每一個0開始處理,先把0替換成1,而後再用 DFS 來找全部相連的1的個數,具體如何找就跟以前的島嶼的題目沒啥區別了,這裏就不細講了,參見代碼以下:blog
解法一:leetcode
class Solution { public: int largestIsland(vector<vector<int>>& grid) { int m = grid.size(), n = grid[0].size(), res = 0; bool hasZero = false; for (int i = 0; i < m; ++i) { for (int j = 0; j < n; ++j) { if (grid[i][j] == 1) continue; grid[i][j] = 1; vector<vector<bool>> visited(m, vector<bool>(n)); res = max(res, helper(grid, i, j, visited)); if (res == m * n) return res; grid[i][j] = 0; hasZero = true; } } return hasZero ? res : m * n; } int helper(vector<vector<int>>& grid, int i, int j, vector<vector<bool>>& visited) { int m = grid.size(), n = grid[0].size(); if (i < 0 || i >= m || j < 0 || j >= n || grid[i][j] == 0 || visited[i][j]) return 0; visited[i][j] = true; return 1 + helper(grid, i - 1, j , visited) + helper(grid, i + 1, j , visited) + helper(grid, i, j - 1, visited) + helper(grid, i, j + 1, visited); } };
固然咱們也能夠用 BFS 來找全部相連的1的個數,整個的解題思路和上面的解法並無什麼不一樣,並不難理解,這可能就是論壇上會有人質疑這道題不該該標爲 Hard 的緣由,參見代碼以下:get
解法二:同步
class Solution { public: int largestIsland(vector<vector<int>>& grid) { int m = grid.size(), n = grid[0].size(), res = 0; bool hasZero = false; vector<int> dirX{0, -1, 0, 1}, dirY{-1, 0, 1, 0}; for (int i = 0; i < m; ++i) { for (int j = 0; j < n; ++j) { if (grid[i][j] == 1) continue; vector<vector<bool>> visited(m, vector<bool>(n)); queue<int> q{{i * n + j}}; int sum = 0; while (!q.empty()) { int t = q.front(); q.pop(); ++sum; for (int k = 0; k < 4; ++k) { int x = t / n + dirX[k], y = t % n + dirY[k]; if (x < 0 || x >= m || y < 0 || y >= n || grid[x][y] == 0 || visited[x][y]) continue; visited[x][y] = true; q.push(x * n + y); } } res = max(res, sum); if (res == m * n) return res; hasZero = true; } } return hasZero ? res : m * n; } };
Github 同步地址:
https://github.com/grandyang/leetcode/issues/827
相似題目:
參考資料:
https://leetcode.com/problems/making-a-large-island/
https://leetcode.com/problems/making-a-large-island/discuss/313787/Two-java-solutions
https://leetcode.com/problems/making-a-large-island/discuss/127256/DFS-JAVA-AC-CONCISE-SOLUTION