Python/graphs/connected_components.py
Nika Losaberidze 2bbdc3bfe7
Implement connected components algorithm for graphs (#2113)
* Implement connected components algorithm for graphs

* fixup! Format Python code with psf/black push

* Add parameters and return values annotations with Python type hints

* updating DIRECTORY.md

* Add doctests and typehints

* Remove unnecessary comments, change variable names

* fixup! Format Python code with psf/black push

Co-authored-by: github-actions <${GITHUB_ACTOR}@users.noreply.github.com>
2020-06-17 18:15:24 +02:00

74 lines
1.5 KiB
Python

"""
https://en.wikipedia.org/wiki/Component_(graph_theory)
Finding connected components in graph
"""
test_graph_1 = {
0: [1, 2],
1: [0, 3],
2: [0],
3: [1],
4: [5, 6],
5: [4, 6],
6: [4, 5],
}
test_graph_2 = {
0: [1, 2, 3],
1: [0, 3],
2: [0],
3: [0, 1],
4: [],
5: [],
}
def dfs(graph: dict, vert: int, visited: list) -> list:
"""
Use depth first search to find all vertexes
being in the same component as initial vertex
>>> dfs(test_graph_1, 0, 5 * [False])
[0, 1, 3, 2]
>>> dfs(test_graph_2, 0, 6 * [False])
[0, 1, 3, 2]
"""
visited[vert] = True
connected_verts = []
for neighbour in graph[vert]:
if not visited[neighbour]:
connected_verts += dfs(graph, neighbour, visited)
return [vert] + connected_verts
def connected_components(graph: dict) -> list:
"""
This function takes graph as a parameter
and then returns the list of connected components
>>> connected_components(test_graph_1)
[[0, 1, 3, 2], [4, 5, 6]]
>>> connected_components(test_graph_2)
[[0, 1, 3, 2], [4], [5]]
"""
graph_size = len(graph)
visited = graph_size * [False]
components_list = []
for i in range(graph_size):
if not visited[i]:
i_connected = dfs(graph, i, visited)
components_list.append(i_connected)
return components_list
if __name__ == "__main__":
import doctest
doctest.testmod()