LeetCode 题解工作台
统计完全连通分量的数量
给你一个整数 n 。现有一个包含 n 个顶点的 无向 图,顶点按从 0 到 n - 1 编号。给你一个二维整数数组 edges 其中 edges[i] = [a i , b i ] 表示顶点 a i 和 b i 之间存在一条 无向 边。 返回图中 完全连通分量 的数量。 如果在子图中任意两个顶点之间…
4
题型
4
代码语言
3
相关题
当前训练重点
中等 · 图·DFS·traversal
答案摘要
我们先根据题目给定的边建立一个邻接表 ,其中 表示顶点 的邻接点集合。 然后我们从 开始遍历所有顶点,如果当前顶点没有被访问过,我们就从当前顶点开始进行深度优先搜索,统计当前连通分量的顶点数 和边数 。如果 $\frac{x(x-1)}{2} = y$,那么当前连通分量就是完全连通分量,我们将答案加一。
Interview AiBoxInterview AiBox 实时 AI 助手,陪你讲清 图·DFS·traversal 题型思路
题目描述
给你一个整数 n 。现有一个包含 n 个顶点的 无向 图,顶点按从 0 到 n - 1 编号。给你一个二维整数数组 edges 其中 edges[i] = [ai, bi] 表示顶点 ai 和 bi 之间存在一条 无向 边。
返回图中 完全连通分量 的数量。
如果在子图中任意两个顶点之间都存在路径,并且子图中没有任何一个顶点与子图外部的顶点共享边,则称其为 连通分量 。
如果连通分量中每对节点之间都存在一条边,则称其为 完全连通分量 。
示例 1:

输入:n = 6, edges = [[0,1],[0,2],[1,2],[3,4]] 输出:3 解释:如上图所示,可以看到此图所有分量都是完全连通分量。
示例 2:

输入:n = 6, edges = [[0,1],[0,2],[1,2],[3,4],[3,5]] 输出:1 解释:包含节点 0、1 和 2 的分量是完全连通分量,因为每对节点之间都存在一条边。 包含节点 3 、4 和 5 的分量不是完全连通分量,因为节点 4 和 5 之间不存在边。 因此,在图中完全连接分量的数量是 1 。
提示:
1 <= n <= 500 <= edges.length <= n * (n - 1) / 2edges[i].length == 20 <= ai, bi <= n - 1ai != bi- 不存在重复的边
解题思路
方法一:DFS
我们先根据题目给定的边建立一个邻接表 ,其中 表示顶点 的邻接点集合。
然后我们从 开始遍历所有顶点,如果当前顶点没有被访问过,我们就从当前顶点开始进行深度优先搜索,统计当前连通分量的顶点数 和边数 。如果 ,那么当前连通分量就是完全连通分量,我们将答案加一。
最后我们返回答案即可。
时间复杂度 ,空间复杂度 。其中 和 分别是顶点数和边数。
class Solution:
def countCompleteComponents(self, n: int, edges: List[List[int]]) -> int:
def dfs(i: int) -> (int, int):
vis[i] = True
x, y = 1, len(g[i])
for j in g[i]:
if not vis[j]:
a, b = dfs(j)
x += a
y += b
return x, y
g = defaultdict(list)
for a, b in edges:
g[a].append(b)
g[b].append(a)
vis = [False] * n
ans = 0
for i in range(n):
if not vis[i]:
a, b = dfs(i)
ans += a * (a - 1) == b
return ans
复杂度分析
| 指标 | 值 |
|---|---|
| 时间 | complexity is O(n + mα(n)) where n is vertices and m is edges, accounting for Union Find operations. Space complexity is O(n) for tracking visited nodes or parent arrays in Union Find. |
| 空间 | O(n) |
面试官常问的追问
外企场景- question_mark
Pay attention if the candidate correctly identifies all connected components using DFS or BFS.
- question_mark
Check if the solution validates completeness efficiently without redundant edge checks.
- question_mark
Watch for proper handling of small graphs and edge cases with single-node components.
常见陷阱
外企场景- error
Forgetting to check every vertex pair in a component can overcount incomplete components.
- error
Revisiting nodes without marking them can lead to infinite loops or wrong component counts.
- error
Assuming all components with edges are complete without verifying full connectivity within each.
进阶变体
外企场景- arrow_right_alt
Count only components with exactly k vertices that are complete.
- arrow_right_alt
Determine the largest complete component in terms of vertex count.
- arrow_right_alt
Handle weighted graphs where completeness requires all edges to exceed a threshold weight.