leetcode day 2 【1905. 统计子岛屿】 BFS/DFS

解题思路

• BFS

找到grid2中的每一座岛屿[暴力搜索整个grid矩阵]，对每座岛屿BFS，过程中check岛屿格子是否在grid1中为岛屿。

class Solution:
    def countSubIslands(self, grid1: List[List[int]], grid2: List[List[int]]) -> int:

        row, col = len(grid2), len(grid2[0])
        def BFS(x:int, y:int) -> int:
            grid2[x][y]  = 0
            q = deque()
            q.append((x, y))
            check = (grid1[x][y] == 1)
            while(q):
                x, y = q.popleft()
                for nx ,ny in [(x-1,y),(x,y-1),(x+1,y),(x,y+1)]:
                    if 0<=nx<row and 0<=ny<col:
                        if grid2[nx][ny] == 1:
                            grid2[nx][ny] = 0
                            q.append((nx,ny))
                            if grid1[nx][ny] == 0:
                                check = False
            return check
            
        count = 0
        for i in range(row):
            for j in range(col):
                if grid2[i][j] == 1:
                    count = count + int(BFS(i,j))
        return count

• DFS

DFS第一种：找到grid2中的每一座岛屿[暴力搜索整个grid矩阵]，对每座岛屿DFS，过程中check岛屿格子是否在grid1中为岛屿。【这里用self.check作为全局变量，DFS中对其修改，无返回值】

class Solution:
    def countSubIslands(self, grid1: List[List[int]], grid2: List[List[int]]) -> int:

        row, col = len(grid2), len(grid2[0])
        self.check = True
        def DFS(x:int, y:int) -> int:
            for nx ,ny in [(x-1,y),(x,y-1),(x+1,y),(x,y+1)]:
                if 0<=nx<row and 0<=ny<col and grid2[nx][ny] == 1:
                    grid2[nx][ny] = 0
                    if grid1[nx][ny] == 0:
                        self.check = False
                    DFS(nx,ny)
            return

        count = 0
        for i in range(row):
            for j in range(col):
                if grid2[i][j] == 1:
                    self.check = (grid1[i][j] == 1)
                    grid2[i][j] = 0
                    DFS(i,j)
                    count = count + int(self.check)
        return count

DFS第二种：【这里用将check作为参数传入DFS中，并作为返回值】。这种方法其实不太好，因为每个递归DFS中都要开辟一个新的叫check的局部变量空间，这是没必要的。

class Solution:
    def countSubIslands(self, grid1: List[List[int]], grid2: List[List[int]]) -> int:

        row, col = len(grid2), len(grid2[0])
        def DFS(x:int, y:int, check:bool) -> int:
            for nx ,ny in [(x-1,y),(x,y-1),(x+1,y),(x,y+1)]:
                if 0<=nx<row and 0<=ny<col and grid2[nx][ny] == 1:
                    grid2[nx][ny] = 0
                    if grid1[nx][ny] == 0:
                        check = False
                    check = DFS(nx,ny,check)
            return check

        count = 0
        for i in range(row):
            for j in range(col):
                if grid2[i][j] == 1:
                    check = (grid1[i][j] == 1)
                    grid2[i][j] = 0
                    checko  = DFS(i,j,check)
                    count = count + int(checko)
        return count

从计算时间和内存消耗来看，还是BFS的效果好。

总结

• 1.python的全局变量、局部变量

很有讲究，在函数内部一般是无法修改全局变量的

• 解决方法是：

【推荐】①用self.check作为可修改的全局变量：详见上面DFS第一种写法
②通过参数传入DFS函数：详见上面DFS第二种写法

---

python中函数的定义,调用，全局变量，局部变量，函数的嵌套使用-初级篇