【发布时间】：2020-12-09 06:03:39
【问题描述】：
在 pg.7 on >this article< 有一个递归算法可以解释给定 co-graph 的 co-tree 的生成。根据这篇文章，我尝试在 python3 中使用库 NetworkX 来实现它，库 NetworkX 用于图结构，库 treelib 用于树数据结构。
为了准确起见，我创建了与示例完全相同的图表（再次在上述文章的第 7 页），并编写了以下函数：
import networkx as nx
from treelib import tree
# Globals
tree = Tree()
CC = 0
def generate_cotree(G,tree_pointer=None,complemented=None):
print("----------------")
global tree
global CC
def _getID():
global CC
CC += 1
return CC
string = "[{0}] Called for graph G = (V,E)\n V = {1} \n E = {2} \n".format(tree_pointer,G.nodes,G.edges)
print(string)
vertexes = list(G.nodes)
edges = list(G.edges)
graph_components = [G.subgraph(c).copy() for c in nx.connected_components(G)]
if len(vertexes) == 1:
'''
G(V,E) where |V| = 0 --> Trivial Cograph
'''
print("...Added leaf node!")
tree.create_node(tag=vertexes[0],parent=tree_pointer)
return 0
elif len(edges) == 0: # Trivial Cograph
'''
G(V,E) where |E| = 0 --> Trivial Cograph
'''
for vertex in vertexes:
tree.create_node(tag=vertex,parent=tree_pointer)
print("...Added leaf node " + str(vertex) + "!")
return 0
else:
if not nx.is_connected(G):
'''
Case A: The G(V,E) is a not connected graph --> Add a 0-node and call the
generate_cotree(g) for every g ε connected_components(G)
'''
print("...Added 0-Node")
id = _getID()
tree.create_node(tag=0,identifier=id,parent=tree_pointer)
tree_ptr = id # Increment the Tree Pointer
for sub_graph in graph_components:
generate_cotree(sub_graph,tree_pointer=tree_ptr)
else:
'''
Case B: The G(V,E) is a connected graph --> Add a 1-node and call the
generate_cotree(g) for every g ε connected_components(~G) where ~G is the complement of G
'''
print("...Added 1-Node")
id = _getID()
tree.create_node(tag=1,identifier=id,parent=tree_pointer)
tree_ptr = id # Increment the Tree Pointer
complement = nx.complement(G) # Get ~G
comp_graph_components = [complement.subgraph(c).copy() for c in nx.connected_components(complement)]
for subgraph in comp_graph_components:
generate_cotree(subgraph,tree_pointer=tree_ptr)
我执行此测试的图 B 的代码如下：
B = nx.Graph()
B.clear()
B.add_node(1)
B.add_node(2)
B.add_node(3)
B.add_node(4)
B.add_node(5)
B.add_node(6)
B.add_node(7)
B.add_node(8)
B.add_edges_from([(2,4),(5,6),(6,7),(5,7),(3,7),(3,6)])
'''
-----
GRAPH:
------
(1) (2)-----(4)
(8) (3)
/ \
/ \
(6)-----(7)
\ /
\ /
(5)
'''
问题是我得到了错误的 co-tree，虽然我可以看到这里出了什么问题，但我不知道如何解决它：
EXPECTED CO-TREE:
(0-node)----------
/ | \ |
(1) (8) \ |
(1-node) (1-node)
/ | / | \
(2) (4) (6) (7) (0-node)
| \
(3) (5)
RESULTING CO-TREE AFTER generate_cotree()
(0-node)----------
/ | \ |
(1) (8) \ |
(1-node) (1-node)
/ | / | \
(2) (4) (6) (7) (1-node)
| \
(3) (5)
有人对这个话题有什么建议吗？
【问题讨论】：
标签： python python-3.x tree networkx graph-theory