def test_top10_central_closeness(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
cc = measures.top10_central_closeness
assert len(cc) == 10
for item in cc:
assert isinstance(item, tuple)
assert len(item) == 2
assert isinstance(item[1], float)
def test_largest_component_measures(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.largest_component_measures.node_count, int)
assert isinstance(measures.largest_component_measures.edge_count, int)
if measures.largest_component_measures.directed:
assert measures.largest_component_measures.strongly_connected
else:
assert measures.largest_component_measures.connected
def test_top10_central_degree(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
dc = measures.top10_central_degree
assert len(dc) == 10
for item in dc:
assert isinstance(item, tuple)
assert len(item) == 2
assert isinstance(item[1], float)
def test_avg_strength(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
if measures.weighted:
if measures.directed:
avg_strength_in, avg_strength_out = measures.avg_strength
assert isinstance(avg_strength_in, float)
assert isinstance(avg_strength_out, float)
else:
assert isinstance(measures.avg_strength, float)
else:
with pytest.raises(MeasureError):
measures.avg_strength
def test_basic_measures(self, graph_path: Path, expected: Any):
measure = GraphMeasures(load_graphml(graph_path))
if measure.directed:
measure.strongly_connected
measure.weakly_connected
with pytest.raises(MeasureError):
measure.connected
else:
measure.connected
with pytest.raises(MeasureError):
measure.weakly_connected
with pytest.raises(MeasureError):
measure.strongly_connected
def test_degree_distribution(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
if measures.directed:
assert isinstance(measures.degree_distribution, tuple)
assert len(measures.degree_distribution) == 2
assert isinstance(measures.degree_distribution[1], dict)
for distribution in measures.degree_distribution:
assert isinstance(distribution, dict)
for degree, count in distribution.items():
assert isinstance(degree, int)
assert isinstance(count, int)
assert issorted(list(distribution.keys()))
else:
assert isinstance(measures.degree_distribution, dict)
for degree, count in measures.degree_distribution.items():
assert isinstance(degree, int)
assert isinstance(count, int)
assert issorted(measures.degree_distribution.keys())
def test_edge_count(self, graph_path: Path, expected: int):
measures = GraphMeasures(load_graphml(graph_path))
assert isclose(measures.edge_count, expected, rel_tol=0.1)
def test_avg_betweenness_centrality(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.avg_betweenness_centrality, float)
def test_component_count(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.component_count, int)
def test_avg_clustering_coefficient(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isclose(measures.avg_clustering_coefficient,
expected,
rel_tol=0.12)
def test_global_efficiency(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.global_efficiency, float)
def test_eccentricity(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.eccentricity, (int, float))
def test_diameter(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.diameter, int)
def test_shortest_path_length(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.shortest_path_length, (int, float))
def test_avg_edge_count(self, graph_path: Path, expected: Any):
measures = GraphMeasures(load_graphml(graph_path))
if measures.directed:
assert trunc(sum(measures.avg_edge_count)) == expected
else:
assert trunc(measures.avg_edge_count) == expected
plt.savefig(results_hist_weight_out.as_posix())
plt.close()
else:
results_hist_weight: Path = results / f"{name}_hist_weight.png"
plt.xlabel('Snaga')
plt.xlabel('Frekvencija')
weights = tuple(edge[2]['weight']
for edge in measures.graph.edges().data())
plt.hist(weights, bins=HISTOGRAM_BINS)
plt.savefig(results_hist_weight.as_posix())
plt.close()
print(f"{datetime.now()} - Graph {name}: FINISHED - histograms")
stat_functions = (write_basic, write_histograms, write_centrality,
write_modularity)
if __name__ == '__main__':
'''
path = graph_paths[4]
p = Process(target=write_modularity, args=(GraphMeasures(load_graphml(path)), path.stem))
p.start()
'''
for graph_id in (0, 1, 2, 3, 4):
path = graph_paths[graph_id]
graph_measures = GraphMeasures(load_graphml(path))
for func in stat_functions:
process = Process(target=func, args=(graph_measures, path.stem))
process.start()
def test_degree_assortativity(self, graph_path: Path, expected):
measures = GraphMeasures(load_graphml(graph_path))
assert isinstance(measures.degree_assortativity, float)