def testAsMirror(self):
name = sys._getframe().f_code.co_name
lst = [[1, 2], [2, 2], [2, 3], [2, 3], [3, 2]]
d = convert.from_edgelist(name, lst, directed=False, overide=True)
print (d.get_edges())
self.assertTrue(not d.is_edge_mirrored)
self.assertEqual(3, d.num_node)
self.assertEqual(3, d.num_edge)
edges = d.get_edges()
assert edges['src'].dtype == np.int
assert edges['dest'].dtype == np.int
self.assertTrue(not load_local(name).is_edge_mirrored)
newname = name + "_new"
d = d.as_mirror_edges(newname=newname, overide=True)
edges = d.get_edges()
print (d.get_edges())
self.assertTrue(d.is_edge_mirrored)
assert edges['src'].dtype == np.int
assert edges['dest'].dtype == np.int
self.assertEqual(3, d.num_node)
self.assertEqual(5, d.num_edge)
self.assertTrue(load_local(newname).is_edge_mirrored)
def testBug6(self):
alg = 'oslom_Infohiermap'
dataname = 'EVAL2_LFR_ud_wu_N1024_mu0.3'
clu = clustering.load_result(dataname, alg)
data = dataset.load_local(dataname)
a = calc_clu_metrics(data, clu)
print(a)
self.assertTrue(not np.isnan(a['cluster_clustering_coefficient']))
def testBug5(self):
alg = 'cdc_MSCD_LFK2'
dataname = 'EVAL2_LFR_ud_wu_N1024_mu0.3'
clu = clustering.load_result(dataname, alg)
data = dataset.load_local(dataname)
calc_clu_metrics(data, clu)
gt = list(data.get_ground_truth().values())[0]
a = calc_clu_gt_attrib(gt, clu)
print(a)
def testBug4(self):
alg = 'scan_pScan'
dataname = 'SIMPLE_ud_wu_nc128_cz8_in7_it4'
clu = clustering.load_result(dataname, alg)
data = dataset.load_local(dataname)
gt = list(data.get_ground_truth().values())[0]
a = calc_clu_metrics(data, gt)
print(a)
a = calc_clu_metrics(data, clu)
print(a)
a = calc_clu_gt_attrib(gt, clu)
print(a)
def load_result(dataname, runname):
fpath = os.path.join(config.get_result_file_path(dataname, runname),
'result.txt')
with open(fpath, 'rt') as f:
j = json.load(f)
d = dataset.load_local(dataname)
edges = d.get_edges()
nodes = set(edges['src']).union(set(edges['dest']))
result_nodes = set()
for l in j['clusters'].values():
result_nodes.update(l)
missed_nodes = nodes.difference(result_nodes)
j['clusters']['-9999'] = list(missed_nodes)
return Result(j)
def load_sample_dataset(name, overide=False):
if not overide and local_exists(name):
return load_local(name)
else:
path = os.path.join(config.GCT_HOME, 'data', _DATASET_[name])
edges = pd.read_csv(path)
gt = pd.read_csv(path.replace("_edges", '_gt'))
description = ""
directed = False
return convert.from_edgelist(name,
edges,
groundtruth=gt,
directed=directed,
description=description,
overide=overide)
def load_snap_dataset(name, overide=False):
if not overide and local_exists(name):
return load_local(name)
else:
conf = _DATASET_[name]
edges = conf.get_edges()
gt = conf.get_ground_truth()
description = conf.description
weighted = conf.weighted
directed = conf.directed
return Dataset(name,
description=description,
groundtruthObj=gt,
edgesObj=edges,
directed=directed,
weighted=weighted,
overide=overide)
def setUp(self):
overide = False
def f(name):
overide = False
return random_dataset.generate_ovp_LFR(name,
N=128,
k=16,
maxk=16,
muw=0.1,
minc=32,
beta=1,
a=0,
overide=overide)
name = "test_LFR_unw_und"
self.graph_unweighted_undirect = self.get_dataset(name, f)
assert not self.graph_unweighted_undirect.is_directed()
assert not self.graph_unweighted_undirect.is_weighted()
def f2(name):
return random_dataset.generate_ovp_LFR("test_LFR_w_und",
N=128,
k=16,
maxk=16,
muw=0.1,
minc=32,
beta=1,
weighted=True,
a=0,
overide=overide)
name = "test_LFR_w_und"
self.graph_weighted_undirect = self.get_dataset(name, f)
assert not self.graph_weighted_undirect.is_directed()
assert self.graph_weighted_undirect.is_weighted()
name = "test_LFR_w_dir"
if dataset.local_exists(name):
self.graph_weighted_direct = dataset.load_local(name)
else:
self.graph_weighted_direct = random_dataset.generate_ovp_LFR(
name,
N=128,
k=16,
maxk=16,
muw=0.1,
minc=32,
beta=1,
weighted=True,
a=1,
overide=overide)
assert self.graph_weighted_direct.is_directed()
assert self.graph_weighted_direct.is_weighted()
name = "test_LFR_unw_dir"
if dataset.local_exists(name):
self.graph_unweighted_direct = dataset.load_local(name)
else:
self.graph_unweighted_direct = random_dataset.generate_ovp_LFR(
name,
N=128,
k=16,
maxk=16,
muw=0.1,
minc=32,
beta=1,
weighted=False,
a=1,
overide=overide)
assert self.graph_unweighted_direct.is_directed()
assert not self.graph_unweighted_direct.is_weighted()
self.graphs = [
self.graph_unweighted_undirect, self.graph_weighted_undirect,
self.graph_weighted_direct, self.graph_unweighted_direct
]
def get_dataset(self, name, fun):
if dataset.local_exists(name):
return dataset.load_local(name)
else:
return fun(name)
def testBug3(self):
alg = 'karateclub_SCD'
dataname = 'EVAL2_LFR_ud_wu_N1024_mu0.3'
clu = clustering.load_result(dataname, alg)
data = dataset.load_local(dataname)
calc_clu_metrics(data, clu)
def testBug2(self):
alg = 'cdc_MSCD_LFK2'
dataname = 'EVAL2_LFR_ud_wu_N1024_mu0.3'
clu = clustering.load_result(dataname, alg)
data = dataset.load_local(dataname)
calc_clu_metrics(data, clu)
