Improve comments, add doctests for kahns_algorithm_topo.py (#11668)

* Improve comments, add doctests for kahns_algorithm_topo.py

* Improve function docstring

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Rename variables, remove print

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
Hardik Pawar 2024-10-04 14:36:08 +05:30 committed by GitHub
parent 40f65e8150
commit e20b503b24
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194

View File

@ -1,36 +1,61 @@
def topological_sort(graph):
def topological_sort(graph: dict[int, list[int]]) -> list[int] | None:
"""
Kahn's Algorithm is used to find Topological ordering of Directed Acyclic Graph
using BFS
Perform topological sorting of a Directed Acyclic Graph (DAG)
using Kahn's Algorithm via Breadth-First Search (BFS).
Topological sorting is a linear ordering of vertices in a graph such that for
every directed edge u v, vertex u comes before vertex v in the ordering.
Parameters:
graph: Adjacency list representing the directed graph where keys are
vertices, and values are lists of adjacent vertices.
Returns:
The topologically sorted order of vertices if the graph is a DAG.
Returns None if the graph contains a cycle.
Example:
>>> graph = {0: [1, 2], 1: [3], 2: [3], 3: [4, 5], 4: [], 5: []}
>>> topological_sort(graph)
[0, 1, 2, 3, 4, 5]
>>> graph_with_cycle = {0: [1], 1: [2], 2: [0]}
>>> topological_sort(graph_with_cycle)
"""
indegree = [0] * len(graph)
queue = []
topo = []
cnt = 0
topo_order = []
processed_vertices_count = 0
# Calculate the indegree of each vertex
for values in graph.values():
for i in values:
indegree[i] += 1
# Add all vertices with 0 indegree to the queue
for i in range(len(indegree)):
if indegree[i] == 0:
queue.append(i)
# Perform BFS
while queue:
vertex = queue.pop(0)
cnt += 1
topo.append(vertex)
for x in graph[vertex]:
indegree[x] -= 1
if indegree[x] == 0:
queue.append(x)
processed_vertices_count += 1
topo_order.append(vertex)
if cnt != len(graph):
print("Cycle exists")
else:
print(topo)
# Traverse neighbors
for neighbor in graph[vertex]:
indegree[neighbor] -= 1
if indegree[neighbor] == 0:
queue.append(neighbor)
if processed_vertices_count != len(graph):
return None # no topological ordering exists due to cycle
return topo_order # valid topological ordering
# Adjacency List of Graph
graph = {0: [1, 2], 1: [3], 2: [3], 3: [4, 5], 4: [], 5: []}
topological_sort(graph)
if __name__ == "__main__":
import doctest
doctest.testmod()