def build_feature_table(flow_path,
ip_mapper,
feature_table,
get_flow_features=True):
path = flow_path + '.tsv'
print('Computing triangle based features ...')
triangle_stats = dict()
gmetrics_lib.get_triangles(path, triangle_stats)
ip_triangle_stats = map_dict(triangle_stats, ip_mapper)
add_feature_list(ip_triangle_stats, feature_table, True, True)
# in_t out_t through_t cycles out_deg in_deg prank krank inf_s out_s inf_d outf_d
# role1: Modest out triangles
# role2: High out degree, modest in degree
# role3: High in triangles
# role4: High out degree, modest out triangles
# role5: High in triangles, low out triangle
# role6: High in degree
# role7: High out flow size
print('Computing pagerank ...')
pagerank = dict()
gmetrics_lib.get_centrality(path, pagerank)
ip_pagerank = map_dict(pagerank, ip_mapper)
add_feature(ip_pagerank, feature_table, True)
print('# ips = ' + str(len(pagerank)))
print('Computing kcore ranking ...')
kcore_ranking = dict()
kcore_ranking = gmetrics_lib.get_kcore_ranking(path)
ip_kcore_ranking = map_dict(kcore_ranking, ip_mapper)
add_feature(ip_kcore_ranking, feature_table, False, True)
print('# ips [1] = ' + str(len(kcore_ranking)))
print('# ips [2] = ' + str(len(ip_kcore_ranking)))
if get_flow_features == False:
return
print('Computing flow size features ...')
median_in_flow = dict()
median_out_flow = dict()
flow_lib.get_median_flow(flow_path, median_in_flow, median_out_flow, True)
add_feature(median_in_flow, feature_table, False)
add_feature(median_out_flow, feature_table, False)
print('# ips = ' + str(len(median_in_flow)))
print('# ips = ' + str(len(median_out_flow)))
print('Computing flow duration features ...')
median_in_flow_duration = dict()
median_out_flow_duration = dict()
flow_lib.get_median_flow(flow_path, \
median_in_flow_duration, median_out_flow_duration, False)
add_feature(median_in_flow_duration, feature_table, False)
add_feature(median_out_flow_duration, feature_table, False)
print('# ips = ' + str(len(median_in_flow_duration)))
print('# ips = ' + str(len(median_out_flow_duration)))
def build_feature_table(flow_path, ip_mapper, feature_table, get_flow_features=True):
path = flow_path + '.tsv'
print('Computing triangle based features ...')
triangle_stats = dict()
gmetrics_lib.get_triangles(path, triangle_stats)
ip_triangle_stats = map_dict(triangle_stats, ip_mapper)
add_feature_list(ip_triangle_stats, feature_table, True, True)
# in_t out_t through_t cycles out_deg in_deg prank krank inf_s out_s inf_d outf_d
# role1: Modest out triangles
# role2: High out degree, modest in degree
# role3: High in triangles
# role4: High out degree, modest out triangles
# role5: High in triangles, low out triangle
# role6: High in degree
# role7: High out flow size
print('Computing pagerank ...')
pagerank = dict()
gmetrics_lib.get_centrality(path, pagerank)
ip_pagerank = map_dict(pagerank, ip_mapper)
add_feature(ip_pagerank, feature_table, True)
print('# ips = ' + str(len(pagerank)))
print('Computing kcore ranking ...')
kcore_ranking = dict()
kcore_ranking = gmetrics_lib.get_kcore_ranking(path)
ip_kcore_ranking = map_dict(kcore_ranking, ip_mapper)
add_feature(ip_kcore_ranking, feature_table, False, True)
print('# ips [1] = ' + str(len(kcore_ranking)))
print('# ips [2] = ' + str(len(ip_kcore_ranking)))
if get_flow_features == False:
return
print('Computing flow size features ...')
median_in_flow = dict()
median_out_flow = dict()
flow_lib.get_median_flow(flow_path, median_in_flow, median_out_flow, True)
add_feature(median_in_flow, feature_table, False)
add_feature(median_out_flow, feature_table, False)
print('# ips = ' + str(len(median_in_flow)))
print('# ips = ' + str(len(median_out_flow)))
print('Computing flow duration features ...')
median_in_flow_duration = dict()
median_out_flow_duration = dict()
flow_lib.get_median_flow(flow_path, \
median_in_flow_duration, median_out_flow_duration, False)
add_feature(median_in_flow_duration, feature_table, False)
add_feature(median_out_flow_duration, feature_table, False)
print('# ips = ' + str(len(median_in_flow_duration)))
print('# ips = ' + str(len(median_out_flow_duration)))
def test_histogram():
triangle_stats = dict()
path = '/pic/projects/mnms4graphs/iscx/tsv/testbed-11jun-aggr.tsv'
#path = 'test.tsv'
gmetrics_lib.get_triangles(path, triangle_stats)
array = []
fout = open('out-deg.dat', 'w')
for k, v in triangle_stats.items():
# read out degree
array.append(int(v[-1]))
fout.write(str(v[-1]) + '\n')
fout.close()
#array = [1, 2, 3, 4, 5, 6]
#array = [[10,-1], [0, 1], [-1, -1]]
#print(sorted(array))
equi_height_histogram(array, 2)
equi_height_histogram(array, 3)
equi_height_histogram(array, 4)
return
def test_histogram():
triangle_stats = dict()
path = '/pic/projects/mnms4graphs/iscx/tsv/testbed-11jun-aggr.tsv'
#path = 'test.tsv'
gmetrics_lib.get_triangles(path, triangle_stats)
array = []
fout = open('out-deg.dat', 'w')
for k, v in triangle_stats.items():
# read out degree
array.append(int(v[-1]))
fout.write(str(v[-1]) + '\n')
fout.close()
#array = [1, 2, 3, 4, 5, 6]
#array = [[10,-1], [0, 1], [-1, -1]]
#print(sorted(array))
equi_height_histogram(array, 2)
equi_height_histogram(array, 3)
equi_height_histogram(array, 4)
return
