46 lines
1.2 KiB
Python
46 lines
1.2 KiB
Python
import networkx as nx
|
|
from networkx.algorithms.approximation import average_clustering
|
|
|
|
# This approximation has to be be exact in regular graphs
|
|
# with no triangles or with all possible triangles.
|
|
|
|
|
|
def test_petersen():
|
|
# Actual coefficient is 0
|
|
G = nx.petersen_graph()
|
|
assert (average_clustering(G, trials=int(len(G) / 2)) ==
|
|
nx.average_clustering(G))
|
|
|
|
|
|
def test_petersen_seed():
|
|
# Actual coefficient is 0
|
|
G = nx.petersen_graph()
|
|
assert (average_clustering(G, trials=int(len(G) / 2), seed=1) ==
|
|
nx.average_clustering(G))
|
|
|
|
|
|
def test_tetrahedral():
|
|
# Actual coefficient is 1
|
|
G = nx.tetrahedral_graph()
|
|
assert (average_clustering(G, trials=int(len(G) / 2)) ==
|
|
nx.average_clustering(G))
|
|
|
|
|
|
def test_dodecahedral():
|
|
# Actual coefficient is 0
|
|
G = nx.dodecahedral_graph()
|
|
assert (average_clustering(G, trials=int(len(G) / 2)) ==
|
|
nx.average_clustering(G))
|
|
|
|
|
|
def test_empty():
|
|
G = nx.empty_graph(5)
|
|
assert average_clustering(G, trials=int(len(G) / 2)) == 0
|
|
|
|
|
|
def test_complete():
|
|
G = nx.complete_graph(5)
|
|
assert average_clustering(G, trials=int(len(G) / 2)) == 1
|
|
G = nx.complete_graph(7)
|
|
assert average_clustering(G, trials=int(len(G) / 2)) == 1
|