def run_experiments(data, ground_truth, network_num):
all_res = []
# Timerank with one connection - default q
mutu4 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='one',
clusters=len(ground_truth),
default_q=True)
all_res.append(
evaluate.get_results(ground_truth,
mutu4.dynamic_coms,
"Timerank with one connection - default q",
mutu4.tfs,
eval="dynamic",
duration=mutu4.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu4.dynamic_coms,
"Timerank with one connection - default q",
mutu4.tfs,
eval="sets",
duration=mutu4.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu4.dynamic_coms,
"Timerank with one connection - default q",
mutu4.tfs,
eval="per_tf",
duration=mutu4.duration))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
# Timerank with all connections - default q
mutu5 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='all',
clusters=len(ground_truth),
default_q=True)
all_res.append(
evaluate.get_results(ground_truth,
mutu5.dynamic_coms,
"Timerank with all connections - default q",
mutu5.tfs,
eval="dynamic",
duration=mutu5.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu5.dynamic_coms,
"Timerank with all connections - default "
"q",
mutu5.tfs,
eval="sets",
duration=mutu5.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu5.dynamic_coms,
"Timerank with all connections - default "
"q",
mutu5.tfs,
eval="per_tf",
duration=mutu5.duration))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
# Timerank with next connection - default q
mutu6 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='next',
clusters=len(ground_truth),
default_q=True)
all_res.append(
evaluate.get_results(ground_truth,
mutu6.dynamic_coms,
"Timerank with next connection - default q",
mutu6.tfs,
eval="dynamic",
duration=mutu6.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu6.dynamic_coms,
"Timerank with next connection - default q",
mutu6.tfs,
eval="sets",
duration=mutu6.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu6.dynamic_coms,
"Timerank with next connection - default q",
mutu6.tfs,
eval="per_tf",
duration=mutu6.duration))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
# NNTF
fact = TensorFact(data.graphs,
num_of_coms=len(ground_truth),
threshold=1e-4,
seeds=10,
overlap=False)
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF",
mutu6.tfs,
eval="dynamic",
duration=fact.duration))
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF",
mutu6.tfs,
eval="sets",
duration=fact.duration))
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF",
mutu6.tfs,
eval="per_tf",
duration=fact.duration))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
with open(results_file, 'a') as f:
f.write("NNTF\n")
f.write("Error: " + str(fact.error) + "Seed: " + str(fact.best_seed) +
"\n")
f.write("A\n")
pprint.pprint(fact.A, stream=f, width=150)
f.write("B\n")
pprint.pprint(fact.B, stream=f, width=150)
f.write("C\n")
pprint.pprint(fact.C, stream=f, width=150)
pprint.pprint(fact.dynamic_coms, stream=f, width=150)
# GED
import sys
sys.path.insert(0, '../GED/')
import preprocessing, Tracker
start_time = time.time()
from ged import GedWrite, ReadGEDResults
ged_data = GedWrite(data)
graphs = preprocessing.getGraphs(ged_data.fileName)
tracker = Tracker.Tracker(graphs)
tracker.compare_communities()
#outfile = 'tmpfiles/ged_results.csv'
outfile = './results/GED-events-handdrawn-' + str(network_num) + '.csv'
with open(outfile, 'w') as f:
for hypergraph in tracker.hypergraphs:
hypergraph.calculateEvents(f)
ged_time = str(datetime.timedelta(seconds=int(time.time() - start_time)))
print "--- %s seconds ---" % (ged_time)
ged = ReadGEDResults.ReadGEDResults(file_coms=ged_data.fileName,
file_output=outfile)
with open(results_file, 'a') as f:
f.write("GED\n")
pprint.pprint(ged.dynamic_coms, stream=f, width=150)
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED",
mutu6.tfs,
eval="dynamic",
duration=ged_time))
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED",
mutu6.tfs,
eval="sets",
duration=ged_time))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
#all_res.append(evaluate.get_results(ground_truth, ged.dynamic_coms, "GED", mutu6.tfs, eval="per_tf"))
# Run Timerank - One connection
mutu1 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='one',
clusters=len(ground_truth),
default_q=False)
all_res.append(
evaluate.get_results(ground_truth,
mutu1.dynamic_coms,
"Timerank with one connection",
mutu1.tfs,
eval="dynamic",
duration=mutu1.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu1.dynamic_coms,
"Timerank with one connection",
mutu1.tfs,
eval="sets",
duration=mutu1.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu1.dynamic_coms,
"Timerank with one connection",
mutu1.tfs,
eval="per_tf",
duration=mutu1.duration))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
muturank_res = OrderedDict()
muturank_res["tf/node"] = ['t' + str(tf) for tf in mutu1.tfs_list]
for i, node in enumerate(mutu1.node_ids):
muturank_res[node] = [
mutu1.p_new[tf * len(mutu1.node_ids) + i]
for tf in range(mutu1.tfs)
]
f = open(results_file, 'a')
f.write("ONE CONNECTION\n")
f.write(
tabulate(muturank_res, headers="keys", tablefmt="fancy_grid").encode(
'utf8') + "\n")
f.write(
tabulate(zip(['t' + str(tf) for tf in mutu1.tfs_list], mutu1.q_new),
headers="keys",
tablefmt="fancy_grid").encode('utf8') + "\n")
f.close()
# Timerank with all connections
mutu2 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='all',
clusters=len(ground_truth),
default_q=False)
all_res.append(
evaluate.get_results(ground_truth,
mutu2.dynamic_coms,
"Timerank with all connections",
mutu2.tfs,
eval="dynamic",
duration=mutu2.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu2.dynamic_coms,
"Timerank with all connections",
mutu2.tfs,
eval="sets",
duration=mutu2.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu2.dynamic_coms,
"Timerank with all connections",
mutu2.tfs,
eval="per_tf",
duration=mutu2.duration))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
muturank_res = OrderedDict()
muturank_res["tf/node"] = ['t' + str(tf) for tf in mutu2.tfs_list]
for i, node in enumerate(mutu2.node_ids):
muturank_res[node] = [
mutu2.p_new[tf * len(mutu2.node_ids) + i]
for tf in range(mutu2.tfs)
]
f = open(results_file, 'a')
f.write("ALL CONNECTIONS\n")
f.write(
tabulate(muturank_res, headers="keys", tablefmt="fancy_grid").encode(
'utf8') + "\n")
f.write(
tabulate(zip(['t' + str(tf) for tf in mutu2.tfs_list], mutu2.q_new),
headers="keys",
tablefmt="fancy_grid").encode('utf8') + "\n")
f.close()
# Timerank with next connection
mutu3 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='next',
clusters=len(ground_truth),
default_q=False)
all_res.append(
evaluate.get_results(ground_truth,
mutu3.dynamic_coms,
"Timerank with next connection",
mutu3.tfs,
eval="dynamic",
duration=mutu3.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu3.dynamic_coms,
"Timerank with next connection",
mutu3.tfs,
eval="sets",
duration=mutu3.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu3.dynamic_coms,
"Timerank with next connection",
mutu3.tfs,
eval="per_tf",
duration=mutu3.duration))
f = open(results_file, 'a')
f.write(
tabulate(all_res, headers="keys", tablefmt="fancy_grid").encode('utf8')
+ "\n")
f.close()
muturank_res = OrderedDict()
muturank_res["tf/node"] = ['t' + str(tf) for tf in mutu3.tfs_list]
for i, node in enumerate(mutu3.node_ids):
muturank_res[node] = [
mutu3.p_new[tf * len(mutu3.node_ids) + i]
for tf in range(mutu3.tfs)
]
f = open(results_file, 'a')
f.write("NEXT CONNECTION\n")
f.write(
tabulate(muturank_res, headers="keys", tablefmt="fancy_grid").encode(
'utf8') + "\n")
f.write(
tabulate(zip(['t' + str(tf) for tf in mutu3.tfs_list], mutu3.q_new),
headers="keys",
tablefmt="fancy_grid").encode('utf8') + "\n")
f.write("GROUND TRUTH\n")
pprint.pprint(ground_truth, stream=f, width=150)
f.write("ONE CONNECTION\n")
pprint.pprint(mutu1.dynamic_coms, stream=f, width=150)
f.write("ALL CONNECTIONS\n")
pprint.pprint(mutu2.dynamic_coms, stream=f, width=150)
f.write("NEXT CONNECTION\n")
pprint.pprint(mutu3.dynamic_coms, stream=f, width=150)
f.close()
return all_res
def run_experiments(data, ground_truth, network_num):
times = []
all_res = []
# Timerank with one connection - default q
start_time = time.time()
mutu4 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='one',
clusters=len(ground_truth),
default_q=True)
all_res.append(
evaluate.get_results(ground_truth,
mutu4.dynamic_coms,
"Timerank-STC-Uni",
mutu4.tfs,
eval="dynamic",
duration=mutu4.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu4.dynamic_coms,
"Timerank-STC-Uni",
mutu4.tfs,
eval="sets",
duration=mutu4.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu4.dynamic_coms,
"Timerank-STC-Uni",
mutu4.tfs,
eval="per_tf",
duration=mutu4.duration))
duration = time.time() - start_time
print("Timerank with one connection - default q: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
# Timerank with all connections - default q
start_time = time.time()
mutu5 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='all',
clusters=len(ground_truth),
default_q=True)
all_res.append(
evaluate.get_results(ground_truth,
mutu5.dynamic_coms,
"Timerank-AOC-Uni",
mutu5.tfs,
eval="dynamic",
duration=mutu5.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu5.dynamic_coms,
"Timerank-AOC-Uni",
mutu5.tfs,
eval="sets",
duration=mutu5.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu5.dynamic_coms,
"Timerank-AOC-Uni",
mutu5.tfs,
eval="per_tf",
duration=mutu5.duration))
duration = time.time() - start_time
print("Timerank with all connection - default q: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
# Timerank with next connection - default q
start_time = time.time()
mutu6 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='next',
clusters=len(ground_truth),
default_q=True)
all_res.append(
evaluate.get_results(ground_truth,
mutu6.dynamic_coms,
"Timerank-NOC-Uni",
mutu6.tfs,
eval="dynamic",
duration=mutu6.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu6.dynamic_coms,
"Timerank-NOC-Uni",
mutu6.tfs,
eval="sets",
duration=mutu6.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu6.dynamic_coms,
"Timerank-NOC-Uni",
mutu6.tfs,
eval="per_tf",
duration=mutu6.duration))
duration = time.time() - start_time
print("Timerank with next connection - default q: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
# Run Timerank - One connection
start_time = time.time()
mutu1 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='one',
clusters=len(ground_truth),
default_q=False)
all_res.append(
evaluate.get_results(ground_truth,
mutu1.dynamic_coms,
"Timerank-STC",
mutu1.tfs,
eval="dynamic",
duration=mutu1.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu1.dynamic_coms,
"Timerank-STC",
mutu1.tfs,
eval="sets",
duration=mutu1.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu1.dynamic_coms,
"Timerank-STC",
mutu1.tfs,
eval="per_tf",
duration=mutu1.duration))
duration = time.time() - start_time
print("Timerank with one connection: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
muturank_res = OrderedDict()
muturank_res["tf/node"] = ['t' + str(tf) for tf in mutu1.tfs_list]
for i, node in enumerate(mutu1.node_ids):
muturank_res[node] = [
mutu1.p_new[tf * len(mutu1.node_ids) + i]
for tf in range(mutu1.tfs)
]
f = open('results_hand.txt', 'a')
f.write("ONE CONNECTION\n")
f.write(tabulate(muturank_res, headers="keys", tablefmt="grid") + "\n")
f.write(
tabulate(zip(['t' + str(tf) for tf in mutu1.tfs_list], mutu1.q_new),
headers="keys",
tablefmt="grid") + "\n")
f.close()
# Timerank with all connections
start_time = time.time()
mutu2 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='all',
clusters=len(ground_truth),
default_q=False)
all_res.append(
evaluate.get_results(ground_truth,
mutu2.dynamic_coms,
"Timerank-AOC",
mutu2.tfs,
eval="dynamic",
duration=mutu2.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu2.dynamic_coms,
"Timerank-AOC",
mutu2.tfs,
eval="sets",
duration=mutu2.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu2.dynamic_coms,
"Timerank-AOC",
mutu2.tfs,
eval="per_tf",
duration=mutu2.duration))
duration = time.time() - start_time
print("Timerank with all connection: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
muturank_res = OrderedDict()
muturank_res["tf/node"] = ['t' + str(tf) for tf in mutu2.tfs_list]
for i, node in enumerate(mutu2.node_ids):
muturank_res[node] = [
mutu2.p_new[tf * len(mutu2.node_ids) + i]
for tf in range(mutu2.tfs)
]
f = open('results_hand.txt', 'a')
f.write("ALL CONNECTIONS\n")
f.write(tabulate(muturank_res, headers="keys", tablefmt="grid") + "\n")
f.write(
tabulate(zip(['t' + str(tf) for tf in mutu2.tfs_list], mutu2.q_new),
headers="keys",
tablefmt="grid") + "\n")
f.close()
# Timerank with next connection
start_time = time.time()
mutu3 = Muturank_new(data.graphs,
threshold=1e-6,
alpha=0.85,
beta=0.85,
connection='next',
clusters=len(ground_truth),
default_q=False)
all_res.append(
evaluate.get_results(ground_truth,
mutu3.dynamic_coms,
"Timerank-NOC",
mutu3.tfs,
eval="dynamic",
duration=mutu3.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu3.dynamic_coms,
"Timerank-NOC",
mutu3.tfs,
eval="sets",
duration=mutu3.duration))
all_res.append(
evaluate.get_results(ground_truth,
mutu3.dynamic_coms,
"Timerank-NOC",
mutu3.tfs,
eval="per_tf",
duration=mutu3.duration))
duration = time.time() - start_time
print("Timerank with next connection: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
muturank_res = OrderedDict()
muturank_res["tf/node"] = ['t' + str(tf) for tf in mutu3.tfs_list]
for i, node in enumerate(mutu3.node_ids):
muturank_res[node] = [
mutu3.p_new[tf * len(mutu3.node_ids) + i]
for tf in range(mutu3.tfs)
]
f = open('results_hand.txt', 'a')
f.write("NEXT CONNECTION\n")
f.write(tabulate(muturank_res, headers="keys", tablefmt="grid") + "\n")
f.write(
tabulate(zip(['t' + str(tf) for tf in mutu3.tfs_list], mutu3.q_new),
headers="keys",
tablefmt="grid") + "\n")
f.write("GROUND TRUTH\n")
pprint.pprint(ground_truth, stream=f, width=150)
f.write("ONE CONNECTION\n")
pprint.pprint(mutu1.dynamic_coms, stream=f, width=150)
f.write("ALL CONNECTIONS\n")
pprint.pprint(mutu2.dynamic_coms, stream=f, width=150)
f.write("NEXT CONNECTION\n")
pprint.pprint(mutu3.dynamic_coms, stream=f, width=150)
f.close()
# NNTF
start_time = time.time()
fact = TensorFact(data.graphs,
num_of_coms=len(ground_truth),
threshold=1e-4,
seeds=1,
overlap=False)
fact_dur = time.time() - start_time
fact_dur = "%d:%d" % (fact_dur // 60, fact_dur % 60)
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF",
mutu6.tfs,
eval="dynamic",
duration=fact_dur))
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF",
mutu6.tfs,
eval="sets",
duration=fact_dur))
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF",
mutu6.tfs,
eval="per_tf",
duration=fact_dur))
duration = time.time() - start_time
print("NNTF: TIME: %d min, %d sec" % (duration // 60, duration % 60))
times.append(duration)
with open('results_hand.txt', 'a') as f:
f.write("NNTF\n")
f.write("Error: " + str(fact.error) + "Seed: " + str(fact.best_seed) +
"\n")
f.write("A\n")
pprint.pprint(fact.A, stream=f, width=150)
f.write("B\n")
pprint.pprint(fact.B, stream=f, width=150)
f.write("C\n")
pprint.pprint(fact.C, stream=f, width=150)
pprint.pprint(fact.dynamic_coms, stream=f, width=150)
# NNTF-Timerank tensor
new_graphs = {}
for i, A in mutu1.a.items():
new_graphs[i] = nx.from_scipy_sparse_matrix(A)
start_time = time.time()
fact2 = TensorFact(new_graphs,
num_of_coms=len(ground_truth),
threshold=1e-4,
seeds=1,
overlap=False,
original_graphs=data.graphs)
fact_dur = time.time() - start_time
fact_dur = "%d:%d" % (fact_dur // 60, fact_dur % 60)
all_res.append(
evaluate.get_results(ground_truth,
fact2.dynamic_coms,
"NNTF-Timerank tensor",
mutu6.tfs,
eval="dynamic",
duration=fact_dur))
all_res.append(
evaluate.get_results(ground_truth,
fact2.dynamic_coms,
"NNTF-Timerank tensor",
mutu6.tfs,
eval="sets",
duration=fact_dur))
all_res.append(
evaluate.get_results(ground_truth,
fact2.dynamic_coms,
"NNTF-Timerank tensor",
mutu6.tfs,
eval="per_tf",
duration=fact_dur))
duration = time.time() - start_time
print("NNTF-Timerank tensor: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
with open('results_hand.txt', 'a') as f:
f.write("NNTF\n")
f.write("Error: " + str(fact2.error) + "Seed: " +
str(fact2.best_seed) + "\n")
f.write("A\n")
pprint.pprint(fact2.A, stream=f, width=150)
f.write("B\n")
pprint.pprint(fact2.B, stream=f, width=150)
f.write("C\n")
pprint.pprint(fact2.C, stream=f, width=150)
pprint.pprint(fact2.dynamic_coms, stream=f, width=150)
# NNTF-Overlap
start_time = time.time()
fact = TensorFact(data.graphs,
num_of_coms=len(ground_truth),
threshold=1e-4,
seeds=1,
overlap=True)
fact_dur = time.time() - start_time
fact_dur = "%d:%d" % (fact_dur // 60, fact_dur % 60)
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF-Overlap",
mutu6.tfs,
eval="dynamic",
duration=fact_dur))
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF-Overlap",
mutu6.tfs,
eval="sets",
duration=fact_dur))
all_res.append(
evaluate.get_results(ground_truth,
fact.dynamic_coms,
"NNTF-Overlap",
mutu6.tfs,
eval="per_tf",
duration=fact_dur))
duration = time.time() - start_time
print("NNTF-Overlap: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
with open('results_hand.txt', 'a') as f:
f.write("NNTF-Overlap\n")
f.write("Error: " + str(fact.error) + "Seed: " + str(fact.best_seed) +
"\n")
f.write("A\n")
pprint.pprint(fact.A, stream=f, width=150)
f.write("B\n")
pprint.pprint(fact.B, stream=f, width=150)
f.write("C\n")
pprint.pprint(fact.C, stream=f, width=150)
pprint.pprint(fact.dynamic_coms, stream=f, width=150)
# GED
import sys
sys.path.insert(0, '../GED/')
start_time = time.time()
from ged import GedWrite, ReadGEDResults
ged_data = GedWrite(data)
graphs = preprocessing.getGraphs(ged_data.fileName)
tracker = Tracker.Tracker(graphs)
tracker.compare_communities()
if not os.path.exists('results'):
os.makedirs('results')
outfile = os.path.join('results',
'GED-events-handdrawn' + str(network_num) + '.csv')
with open(outfile, 'w+') as f:
for hypergraph in tracker.hypergraphs:
hypergraph.calculateEvents(f)
ged = ReadGEDResults.ReadGEDResults(file_coms=ged_data.fileName,
file_output=outfile)
ged_dur = time.time() - start_time
ged_dur = "%d:%d" % (ged_dur // 60, ged_dur % 60)
with open('results_hand.txt', 'a') as f:
f.write("GED\n")
pprint.pprint(ged.dynamic_coms, stream=f, width=150)
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED-T",
mutu6.tfs,
eval="dynamic",
duration=ged_dur))
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED-T",
mutu6.tfs,
eval="sets",
duration=ged_dur))
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED-T",
mutu6.tfs,
eval="per_tf",
duration=ged_dur))
duration = time.time() - start_time
print("GED-T: TIME: %d min, %d sec" % (duration // 60, duration % 60))
times.append(duration)
# GED with timerank communities
# GED
import sys
sys.path.insert(0, '../GED/')
start_time = time.time()
from ged import GedWrite, ReadGEDResults
ged_data = GedWrite(
Data(mutu1.comms, data.graphs, len(graphs), len(mutu1.dynamic_coms),
mutu1.dynamic_coms))
graphs = preprocessing.getGraphs(ged_data.fileName)
tracker = Tracker.Tracker(graphs)
tracker.compare_communities()
outfile = './results/GED-events-handdrawn-' + str(network_num) + '.csv'
with open(outfile, 'w') as f:
for hypergraph in tracker.hypergraphs:
hypergraph.calculateEvents(f)
ged = ReadGEDResults.ReadGEDResults(file_coms=ged_data.fileName,
file_output=outfile)
ged_dur = time.time() - start_time
ged_dur = "%d:%d" % (ged_dur // 60, ged_dur % 60)
with open('results_hand.txt', 'a') as f:
f.write("GED\n")
pprint.pprint(ged.dynamic_coms, stream=f, width=150)
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED - with Timerank comms",
mutu6.tfs,
eval="dynamic",
duration=ged_dur))
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED - with Timerank comms",
mutu6.tfs,
eval="sets",
duration=ged_dur))
all_res.append(
evaluate.get_results(ground_truth,
ged.dynamic_coms,
"GED - with Timerank comms",
mutu6.tfs,
eval="per_tf",
duration=ged_dur))
duration = time.time() - start_time
print("GED - with Timerank comms: TIME: %d min, %d sec" %
(duration // 60, duration % 60))
times.append(duration)
print("TOTAL TIME: %d min, %d sec" % (sum(times) // 60, sum(times) % 60))
return all_res