Topological Sorting
update Jul 18, 2017 11:11
Given an directed graph, a topological order of the graph nodes is defined as follow:
For each directed edge A -> B in graph, A must before B in the order list. The first node in the order can be any node in the graph with no nodes direct to it. Find any topological order for the given graph.
Notice
You can assume that there is at least one topological order in the graph.
Example
For graph as follow:
The topological order can be:
[0, 1, 2, 3, 4, 5]
[0, 2, 3, 1, 5, 4]
...Challenge Can you do it in both BFS and DFS?
Basic Idea:
使用Kahn's Algorithm, 因为直接dfs会由于递归栈的关系存在很多隐患，而Kahn算法可以使用BFS。
具体地，说一下大体思路：
首先统计每个node的indegree，存在map或者dict里面。
把indegree为0的node加入res和queue，开始bfs
在bfs的过程中，每次把node的neighbors的indegree都减一（相当于去掉这个node以及其发出的路径），把indegree为0的neighbor加入res和queue，继续bfs。
类似地，dfs的方法就是在遇到neighbor的indegree减一之后变为0时，马上跟进进行dfs。
Follow up: 如何判断能否进行 topological sort 呢（即是否有环）？
如果bfs结束之后，仍有indegree > 0的node，则说明至少有一个环。
如果采用下面python dfs的类似方法也是一样，如果仍存在大于0的indegree的node，则有环。
(待续)
Java Code:
/**
* Definition for Directed graph.
* class DirectedGraphNode {
* int label;
* ArrayList<DirectedGraphNode> neighbors;
* DirectedGraphNode(int x) { label = x; neighbors = new ArrayList<DirectedGraphNode>(); }
* };
*/
public class Solution {
/**
* @param graph: A list of Directed graph node
* @return: Any topological order for the given graph.
*/
// BFS solution
public ArrayList<DirectedGraphNode> topSort(ArrayList<DirectedGraphNode> graph) {
// 先用一个hashmap count indegree
Map<DirectedGraphNode, Integer> count = new HashMap<>();
for (DirectedGraphNode node : graph) {
count.put(node, 0);
}
for (DirectedGraphNode node : graph) {
for (DirectedGraphNode neighbor : node.neighbors) {
count.put(neighbor, count.get(neighbor) + 1);
}
}
// 选择indegree为0的所有node，加入res和queue，开始bfs
ArrayList<DirectedGraphNode> res = new ArrayList<>();
Deque<DirectedGraphNode> queue = new LinkedList<>();
for (DirectedGraphNode node : graph) {
if (count.get(node) == 0) {
queue.addFirst(node);
res.add(node);
}
}
while (! queue.isEmpty()) {
DirectedGraphNode node = queue.removeLast();
for (DirectedGraphNode neighbor : node.neighbors) {
count.put(neighbor, count.get(neighbor) - 1);
if (count.get(neighbor) == 0) {
res.add(neighbor);
queue.addFirst(neighbor);
}
}
}
return res;
}
}Python Code:
# Definition for a Directed graph node
# class DirectedGraphNode:
# def __init__(self, x):
# self.label = x
# self.neighbors = []
class Solution:
"""
@param graph: A list of Directed graph node
@return: A list of graph nodes in topological order.
"""
def topSort(self, graph):
def dfs(count, res, node):
#这里比较关键，如果dfs之后，dfs入口的indegree count仍然为0，则有可能被
#最外层for loop计算两次，所以这里手动将dfs过的点的indegree设为-1，既可避免重复
count[node] = -1
res.append(node)
for neighbor in node.neighbors:
count[neighbor] -= 1
if count[neighbor] == 0:
dfs(count, res, neighbor)
# 先用一个dict统计每个node的indegree
count = collections.defaultdict(int)
for node in graph:
for neighbor in node.neighbors:
count[neighbor] += 1
# 然后对所有indegree为0的node进行dfs，每次把相应node所有neighbor的indgree -= 1，
# 为0的进入下一轮dfs
res = []
for node in graph:
if count[node] == 0:
dfs(count, res, node)
return resupdate 2018-05-19 15:36:48
C++ DFS Solution
因为topological sort 顺序中考前的node一定最后结束dfs，所以只要进行一次dfs，将每个结束dfs的node顺序插入list尾部，最后将list反向再返回就好了。
public class Solution {
/*
* @param graph: A list of Directed graph node
* @return: Any topological order for the given graph.
*/
public ArrayList<DirectedGraphNode> topSort(ArrayList<DirectedGraphNode> graph) {
ArrayList<DirectedGraphNode> res = new ArrayList<>();
Set<Integer> visited = new HashSet<>();
for (DirectedGraphNode node : graph) {
dfs(node, res, visited);
}
Collections.reverse(res);
return res;
}
private void dfs(DirectedGraphNode node, List<DirectedGraphNode> res, Set<Integer> visited) {
if (visited.contains(node.label)) return;
visited.add(node.label);
for (DirectedGraphNode neighbor : node.neighbors) {
dfs(neighbor, res, visited);
}
res.add(node);
}
}C++ Kahn's Algorithm Solution
class Solution {
public:
/*
* @param graph: A list of Directed graph node
* @return: Any topological order for the given graph.
*/
vector<DirectedGraphNode*> topSort(vector<DirectedGraphNode*>& graph) {
// 统计每个node的 indegree
unordered_map<DirectedGraphNode*, int> _count;
for (auto* node : graph) {
_count[node];
for (auto* neighbor : node->neighbors) {
++_count[neighbor];
}
}
// 将 indegree==0 的 node 放入 queue
deque<DirectedGraphNode*> _queue;
for (auto _pair : _count) {
if (_pair.second == 0) _queue.push_back(_pair.first);
}
// BFS，每次将当前indegree为0的node的所有neighbors indegree -1，如果为0，
// 加入 queue
vector<DirectedGraphNode*> res;
while (! _queue.empty()) {
auto* node = _queue.front();
_queue.pop_front();
for (auto* neighbor : node->neighbors) {
if (--_count[neighbor] == 0) {
_queue.push_back(neighbor);
}
}
res.push_back(node);
}
return res;
}
};