def get_hits_venues():
mapping = snap.TStrIntSH()
t0 = time()
file_output_1 = open("paper_venues_hits_hub.txt", 'w')
file_output_2 = open("paper_venues_hits_auth.txt", 'w')
G0 = snap.LoadEdgeListStr(snap.PNGraph, "paperid_venueid_ref.txt", 0, 1,
mapping)
NIdHubH = snap.TIntFltH()
NIdAuthH = snap.TIntFltH()
snap.GetHits(G0, NIdHubH, NIdAuthH, 1000)
print("HITS time:", round(time() - t0, 3), "s")
for item in NIdHubH:
file_output_1.write(
str(mapping.GetKey(item)) + "," + str(NIdHubH[item]) + '\n')
for item in NIdAuthH:
file_output_2.write(
str(mapping.GetKey(item)) + "," + str(NIdAuthH[item]) + '\n')
# convert input string to node id
# NodeId = mapping.GetKeyId("814DF491")
# convert node id to input string
# NodeName = mapping.GetKey(NodeId)
# print "name", NodeName
# print "id ", NodeId
print("finish hits!")
file_output_1.close()
file_output_2.close()
def load_graph(self, name):
'''
Helper function to load graphs.
Check that the respective .txt files are in the same folder as this script;
if not, change the paths below as required.
'''
mapping = snap.TStrIntSH()
G = snap.LoadEdgeListStr(snap.PNGraph, self.graphName, 0, 1, mapping)
return G
def hits(graph_filename):
# create graph
name_id_map = snap.TStrIntSH()
graph = snap.LoadEdgeListStr(snap.PNGraph, graph_filename, 0, 1,
name_id_map)
# run HITS algo
id_hub_map = snap.TIntFltH()
id_auth_map = snap.TIntFltH()
snap.GetHits(graph, id_hub_map, id_auth_map, 1000) # iterate 1000 times
return name_id_map, id_hub_map, id_auth_map
def getDataSource(forceReload=False):
dataSourceFileName = "scoring.pkl"
import os.path
if forceReload or not os.path.isfile(dataSourceFileName):
Data_Scraper.load_data()
print('before mapping')
mapping = snap.TStrIntSH()
G = snap.LoadEdgeListStr(snap.PNGraph, "politics_edge_list.txt", 0, 1,
mapping)
rankedCommentsPos = []
rankedCommentsNeg = []
combinedNeg = []
combinedPos = []
#1000000
commentsToTake = 5000
for i in range(1, 5):
rankedCommentData = Snap_Analytics.sort_comments(60, mapping, i)
rankedCommentsPos.append(rankedCommentData[:commentsToTake])
rankedCommentsNeg.append(
rankedCommentData[len(rankedCommentData) -
commentsToTake * 4:len(rankedCommentData)])
combinedPos.extend(rankedCommentData[:commentsToTake / 4])
combinedNeg.extend(rankedCommentData[len(rankedCommentData) -
(commentsToTake * 4) /
4:len(rankedCommentData)])
rankedCommentsPos.append(combinedPos)
rankedCommentsNeg.append(combinedNeg)
f = open(dataSourceFileName, 'wb')
pickle.dump((rankedCommentsPos, rankedCommentsNeg), f)
f.close()
return (rankedCommentsPos, rankedCommentsNeg)
else:
print("loading from file")
f = open(dataSourceFileName, 'rb')
classifier = pickle.load(f)
f.close()
return (classifier[0], classifier[1])
import snap
H = snap.TStrIntSH()
Graph = snap.LoadConnListStr(snap.PNGraph, "data/example-LoadConnListStr.txt",
H)
# get node ID of "A"
print H.GetDat("A")
H = snap.TStrIntSH()
UGraph = snap.LoadConnListStr(snap.PUNGraph,
"data/example-LoadConnListStr.txt", H)
H = snap.TStrIntSH()
Network = snap.LoadConnListStr(snap.PNEANet,
"data/example-LoadConnListStr.txt", H)
def load_graph(file_path):
H = snap.TStrIntSH()
Graph = snap.LoadConnListStr(snap.PNGraph, file_path, H)
print "-----graph loaded"
return Graph, H
def main():
print('begin main')
Data_Scraper.load_data()
#measure_comment_lengths()
# measure_comment_lengths()
# FK_histogram()
print('before mapping')
#for i in xrange(25):
# sample_random_comment(G, mapping)
mapping = snap.TStrIntSH()
G = snap.LoadEdgeListStr(snap.PNGraph, "politics_edge_list.txt", 0, 1, mapping)
root = getRootNode(mapping, G)
# FK_histogram()
# FK_scores = getFKScores()
# """ TRYING TO CREATE A SMALL GRAPH SAMPLE FOR VISUALIZATION """
# f = open("Mini_graph.csv", "w")
# thread_id = random.sample(Data_Scraper.thread_ids, 1)
# newRoot = G.GetNI(mapping.GetKeyId(thread_id)).GetId()
# # newRoot = mapping.GetKeyId(thread_id)
# f.write(str(root.GetId()) + "," + str(newRoot) + "\n")
# sg = makeSubGraph(G, mapping)
# for edge in sg.Edges():
# tup = edge.GetId()
# f.write(str(tup[0]) + "," + str(tup[1]) + "\n")
# f.close()
# print "File written"
# for key,value in FK_scores.iteritems():
# f.write(str(key) + "\t" + str(value) + "\n")
# f.close()
# print "File written"
#stats_vec = comment_statistics(mapping, G)
#print(stats_vec[4])
#getCommentHistogram([G.GetNI(n) for n in root.GetOutEdges()], G)
#output = open('comment_stats.pkl', 'wb')
#pickle.dump(stats_vec,output)
#output.close()
#sort_comments(200, mapping, 1)
# pickle.dump(stats_vec,output)
#output.close()
# print "Nodes: " + str(Data_Scraper.all_comments)
# print "Threads: " + str(len(Data_Scraper.thread_ids))
# print "Root comments: " + str(len(Data_Scraper.root_comments))
#print "delay: " + estimated_delay(144834, mapping, G)
#record_comment_lengths()
assemble_sequence_dict(mapping)
print G.GetNodes()
print G.GetEdges()
nodes.append(node)
links = []
print("looping thru edges")
for i in G.Edges():
edge = {}
edge["source"] = mapping.GetKey(i.GetSrcNId())
edge["target"] = mapping.GetKey(i.GetDstNId())
edge["value"] = 1
links.append(edge)
f = open("graph.json", "w")
f.write(json.dumps({"nodes": nodes, "links": links}))
f.close()
# TODO: mean and standard deviation of FK - scores and word lengths
if __name__ == "__main__":
main()
mapping = snap.TStrIntSH()
G = snap.LoadEdgeListStr(snap.PNGraph, "politics_edge_list.txt", 0, 1, mapping)
# root = getRootNode(mapping, G)
# thread_sizes = {}
# for thread in root.GetOutEdges():
# threadsize = _findDepth(G.GetNI(thread), G)
# thread_sizes[mapping.GetKey(thread)] = threadsize