In a 2D grid of `0`s and `1`s, we change at most one `0` to a `1`.
After, what is the size of the largest island? (An island is a 4-directionally connected group of 1
s).
Example 1:
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:
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:
Input: [[1, 1], [1, 1]]
Output: 4
Explanation: Can\'t change any 0 to 1, only one island with area = 4.
Notes:
1 <= grid.length = grid[0].length <= 50
.0 <= grid[i][j] <= 1
.
这道题在只有0和1的矩阵中用相连的1来表示小岛,现在说是有一个把0变为1的机会,这样原本不相邻的岛就有可能变的相邻了,从而组成一个更大的岛,让求出可能组成的最大的岛屿的面积,也就是相连的1的个数。在 LeetCode 中关于岛屿的题其实也做过许多,比如 [Number of Distinct Islands II](http://www.cnblogs.com/grandyang/p/8542820.html),[Max Area of Island](http://www.cnblogs.com/grandyang/p/7712724.html),[Number of Distinct Islands](http://www.cnblogs.com/grandyang/p/7698778.html),[Number of Islands II](http://www.cnblogs.com/grandyang/p/5190419.html),和 [Number of Islands](http://www.cnblogs.com/grandyang/p/4402656.html)。其实大多题目的本质都是用 DFS 或者 BFS 去遍历所有相连的1,当然这道题也不例外。博主最开始的想法是首先用 DFS 来找出每个岛屿,然后把同一个岛屿的位置坐标都放到同一个 HashSet 中,这样就有了很多 HashSet,然后遍历所有的0的位置,对每个0位置,遍历其周围4个邻居,然后看邻居位置有没有属于岛屿的,有的话就把该岛屿的 HashSet 编号记录下来,遍历完4个邻居后,在把所有的相连的岛屿中的1个数加起来(因为 HashSet 可以直接求出集合中数字的总个数),每次更新结果 res 即可。这种方法是可以通过 OJ 的,速度还比下面展示的两种方法要快,就是代码比较长,没有下面方法的简洁,这里就不贴了。下面的这两种方法其实都是从每个0开始处理,先把0替换成1,然后再用 DFS 来找所有相连的1的个数,具体如何找就跟之前的岛屿的题目没啥区别了,这里就不细讲了,参见代码如下:
解法一:
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);
}
};
当然我们也可以用 BFS 来找所有相连的1的个数,整个的解题思路和上面的解法并没有什么不同,并不难理解,这可能就是论坛上会有人质疑这道题不应该标为 Hard 的原因,参见代码如下:
解法二:
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
类似题目:
参考资料:
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
[LeetCode All in One 题目讲解汇总(持续更新中...)](https://www.cnblogs.com/grandyang/p/4606334.html)