def do_load_dblp(self,
filename,
start_year=2000,
end_year=2004,
conf_file='data/dblp/confs.txt',
coms='components'):
"""
Load dblp dataset between years given
:param filename:
:param start_year:
:param end_year:
:param coms: ground truth communities or conferences 'confs' or 'components'
:return:
"""
if not filename:
filename = "data/dblp/my_dblp_data.json"
dblp = dblp_loader(filename,
start_year=start_year,
end_year=end_year,
conf_file=conf_file,
coms=coms)
self.graphs = dblp.graphs
self.comms = dblp.communities
self.timeFrames = dblp.timeFrames
print "DBLP data have been successfully loaded!"
return
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
if __name__ == "__main__":
from os.path import expanduser
home = expanduser("~")
path_full = home + "/Dropbox/Msc/thesis/data/dblp/1"
results_file = "results_dblp_" + path_full.split("/")[-1] + ".txt"
print results_file
dblp = dblp_loader(path_full + "/my_dblp_data.json",
conf_file=path_full + "/confs.txt",
start_year=2006,
end_year=2010,
coms='comp')
f = open(results_file, 'a')
for item in dblp.conf_list:
f.write("%s\n" % item)
f.close()
# for t in dblp.graphs.keys():
# print nx.number_of_nodes(dblp.graphs[t]),
# new_edges = []
# for conf, graph in dblp.conf_graphs[t].iteritems():
# new_edges.extend(graph.edges())
# print nx.number_of_nodes(nx.Graph(new_edges))
# print(len(set(item for sublist in [g.nodes() for _, g in dblp.graphs.items()] for item in sublist)))
for tf, year in enumerate(dblp.communities.keys()):
print len(dblp.all_nodes[tf]), len(