def main():
random.seed(13) # the seed can affect affect the search path
gfile = sys.argv[1] # data graph file given as GraphML
qfiles = sys.argv[
2:] # list of graphs (GraphML) to query in the data graph
print("---- Graph: %s ----" % (gfile))
st = time.time()
G = igraph.load(gfile, format="graphml")
load_time = time.time() - st
print("---- times ----")
print("load time: %.2f" % load_time)
total_time = 0.0
for qfile in qfiles:
print("---- Query: %s ----" % (qfile))
st = time.time()
Q = igraph.load(qfile, format="graphml")
loadQ_time = time.time() - st
iso_time, num_matches = vf2(G, Q)
query_time = loadQ_time + iso_time
print("query time : %.2f" % query_time)
print("num matches: %d" % (num_matches))
total_time += query_time
print("total query time: %.2f" % (total_time))
print("total time: %.2f" % (total_time + load_time))
print("\n")
def process_file(self, filename):
"""Loads a graph from the given file, runs the clustering
algorithm on it and prints the clusters to the standard
output."""
self.log.info("Processing %s..." % filename)
_, ext = os.path.splitext(filename)
if ext == ".txt":
graph = load(filename, format=self.options.format, directed=False)
else:
graph = load(filename, format=self.options.format)
# If the graph has weights and we want to ignore them, delete them
if self.options.no_weights and "weight" in graph.edge_attributes():
del graph.es["weight"]
# If the graph is directed, we have to make it undirecteed
if graph.is_directed():
self.log.warning("Converting directed graph to undirected.")
if igraph_version < "0.6":
graph.to_undirected(collapse=False)
self.log.warning("Mutual edges will be collapsed into a single edge.")
self.log.warning("Weights will be lost!")
else:
graph.to_undirected(combine_edges="sum")
# Sanity checks
if not graph.is_simple():
self.log.warning("Removing self-loops and collapsing multiple edges...")
if igraph_version < "0.6":
graph.simplify(reduce_attributes="sum")
else:
graph.simplify(combine_edges="sum")
# Set up the "name" attribute properly
if "label" in graph.vertex_attributes():
graph.vs["name"] = graph.vs["label"]
del graph.vs["label"]
elif "name" not in graph.vertex_attributes():
graph.vs["name"] = [str(i) for i in xrange(graph.vcount())]
# Run the algorithm, get the result generator
self.log.info("Calculating clusters, please wait...")
algorithm = HLC(graph, self.options.min_size)
results = algorithm.run(self.options.threshold)
# Print the optimal threshold if we determined it automatically
if self.options.threshold is None:
self.log.info("Threshold = %.6f" % algorithm.last_threshold)
self.log.info("D = %.6f" % algorithm.last_partition_density)
# Print the results
if self.options.output:
outfile = open(self.options.output, "w")
self.log.info("Saving results to %s..." % self.options.output)
else:
outfile = sys.stdout
for community in results:
print("\t".join(graph.vs[community]["name"]), file=outfile)
def main():
G = igraph.load('/Users/ytesfaye/tmp/GRAPHS/flights.graphml')
#snap_home, filename = setup(G)
vc = cesna(G)
print(vc)
def __prepare__(data_dir):
"""
TEMPLATE COMMENT: prepare the data into graphml format.
"""
vertex_filename = os.path.join(data_dir, VERTEX_DATA_FILE)
edge_filename = os.path.join(data_dir, EDGE_DATA_FILE)
g = igraph.load(edge_filename)
vertex_file = open(vertex_filename)
reader = csv.DictReader(vertex_file)
for case in reader:
caseid = int(case['caseid'])
v = g.vs[caseid]
v['caseid'] = case['caseid']
v['usid'] = case['usid']
v['parties'] = case['parties']
v['year'] = case['year']
vertex_file.close()
# Case IDs are 1-indexed, so we delete the 0th vertex as it was extraneous.
g.delete_vertices([0])
graph_file = os.path.join(data_dir, GRAPH_NAME + GRAPH_TYPE)
g.write_graphml(graph_file)
def load_data():
start = time.time()
try:
print("Loading data from /data pickles and hfd5 adj matrices")
f = h5py.File('data/cosponsorship_data.hdf5', 'r')
for chamber in ['house', 'senate']:
for congress in SUPPORTED_CONGRESSES:
adj_matrix_lookup[(chamber, congress)] = np.asarray(f[chamber + str(congress)])
igraph_graph = igraph.load("data/" + chamber + str(congress) + "_igraph.pickle", format="pickle")
igraph_graph_lookup[(chamber, congress, False)] = igraph_graph
nx_graph = nx.read_gpickle("data/" + chamber + str(congress) + "_nx.pickle")
nx_graph_lookup[(chamber, congress, False)] = nx_graph
except IOError as e:
print("Loading data from cosponsorship files")
f = h5py.File("data/cosponsorship_data.hdf5", "w")
for chamber in ['house', 'senate']:
for congress in SUPPORTED_CONGRESSES:
print("Starting %s %s" % (str(congress), chamber))
adj_matrix = load_adjacency_matrices(congress, chamber)
data = f.create_dataset(chamber + str(congress), adj_matrix.shape, dtype='f')
data[0: len(data)] = adj_matrix
# igraph
get_cosponsorship_graph(congress, chamber, False).save("data/" + chamber + str(congress) + "_igraph.pickle", "pickle")
# networkx
nx.write_gpickle(get_cosponsorship_graph_nx(congress, chamber, False), "data/" + chamber + str(congress) + "_nx.pickle")
print("Done with %s %s" % (str(congress), chamber))
print("Data loaded in %d seconds" % (time.time() - start))
def __prepare__(data_dir):
"""
TEMPLATE COMMENT: prepare the data into graphml format.
"""
vertex_filename = os.path.join(data_dir, VERTEX_DATA_FILE)
edge_filename = os.path.join(data_dir, EDGE_DATA_FILE)
g = igraph.load(edge_filename)
vertex_file = open(vertex_filename)
reader = csv.DictReader(vertex_file)
for case in reader:
caseid = int(case['caseid'])
v = g.vs[caseid]
v['caseid'] = case['caseid']
v['usid'] = case['usid']
v['parties'] = case['parties']
v['year'] = case['year']
vertex_file.close()
# Case IDs are 1-indexed, so we delete the 0th vertex as it was extraneous.
g.delete_vertices([0])
graph_file = os.path.join(data_dir, GRAPH_NAME + GRAPH_TYPE)
g.write_graphml(graph_file)
def load(fn):
"""
Load a graph from file with name fn.
Returns a sonet.graph.Graph
"""
return Graph(ig.load(fn))
def run_graph(graph_file, output_file, neighborhood, flag = None):
print "==== New graph ===="
print "Input:", graph_file
print "Output:", output_file
print "Neighborhood:", neighborhood
print ""
print "Loading graph"
print "Start time: %s" % (datetime.now())
g = ig.load(graph_file)
print "End time: %s" % (datetime.now())
gflag = None
if flag == None:
print "Creating flag complex"
print "Start time: %s" % (datetime.now())
gedges = list(g.get_edgelist())
gedgelist = map(list,gedges)
gflag = sh.flag(gedgelist,4)
print "End time: %s" % (datetime.now())
else:
print "Using passed in flag complex"
gflag = flag
print "Finding the local homology"
print "Start time: %s" % (datetime.now())
for n in range(neighborhood+1):
graph = graph_file.split("/")[-1]
ofile = open(output_file, 'w')
ofile.write("Local Homology (neighborhood=%d) of flag complex generated by %s\n" % (n, graph))
locHomTable(gflag, {}, n, ofile)
print "End time: %s" % (datetime.now())
print "\n"
return gflag
def networking(nodes, edges):
#creates graph
net = nx.Graph()
#adds nodes to graph
for n in nodes:
net.add_node(str(n.id1))
for s1 in edges.keys():
for s2 in edges[s1].keys():
#net.add_edge(int(s1.id1),int(s2.id1), length=edges[s1][s2]["cosine"])
net.add_edge(str(s1.id1),
str(s2.id1),
weight=edges[s1][s2]["cosine"])
#prints number of nodes in net
print(net.number_of_nodes())
#exports .graphml and exports .svg to browser
nx.write_graphml(net, "graph.graphml")
nx.write_gml(net, "graph.gml")
graph = igraph.load("graph.gml")
layout = graph.layout("fr")
igraph.Graph.write_svg(graph,
fname="static/graph_image.svg",
layout=layout,
width=4000,
height=4000)
return net
def main():
G = igraph.load('/Users/ytesfaye/tmp/GRAPHS/flights.graphml')
#snap_home, filename = setup(G)
vc = cesna(G)
print(vc)
def load(fn):
"""
Load a graph from file with name fn.
Returns a sonet.graph.Graph
"""
return Graph(ig.load(fn))
def _draw(args):
import igraph
g = igraph.load('_'.join(args.screen_name) + EDG_EXT)
sys.stdout.write('%s Handles\n%s Follow Relationships\n' %
(g.vcount(), g.ecount()))
sys.stdout.write('Avg Shortest Path = %.6f\n' % g.average_path_length())
sys.stdout.write('In-Degree Distribution mean = %.6f, sd = %.6f\n' %
(g.degree_distribution(mode=igraph.IN).mean,
g.degree_distribution(mode=igraph.IN).sd))
sys.stdout.write('Clustering Coefficient = %.6f\n' %
g.transitivity_undirected())
sys.stdout.write('Degree Assortativity = %.6f\n' %
g.assortativity_degree())
width = height = int(750 + g.vcount() * 4.73 - g.vcount()**2 * 1.55e-3)
comp = g.clusters(igraph.STRONG if args.strong else igraph.WEAK)
sys.stdout.write('Cluster Modularity = %.6f\n' % comp.modularity)
gc = comp.giant()
gc_size = gc.vcount()
sys.stdout.write('Fraction Handles in Giant Component = %.2f%%\n' %
(100.0 * gc_size / g.vcount()))
if len(args.screen_name) > 1 and not args.strong:
RGB_dict = dict(zip(SHAPES, _palette(len(SHAPES))))
for v in g.vs:
v['color'] = RGB_dict[v['shape']]
else:
for v in g.vs:
v['color'] = (.8, .8, .8)
if gc.ecount() > 0:
sys.stdout.write('GC Diameter %i (unweighted)\n' % gc.diameter())
gc_vert = comp[comp.sizes().index(gc_size)]
cim = gc.community_infomap(edge_weights=gc.es['weight'],
vertex_weights=gc.vs['lfr'])
RGB_tuples = _palette(max(cim.membership) + 1)
for v in g.vs:
gc_v = gc.vs.select(id_eq=v['id'])
if len(gc_v) > 0:
v['color'] = RGB_tuples[cim.membership[gc_v[0].index]]
g.es['color'] = [g.vs['color'][e.target] for e in g.es]
E = max(g.es['weight'])
g.es['width'] = [(not args.transparent) * max(1, 10 * e['weight'] / E)
for e in g.es]
g.es['arrow_size'] = [max(1, 3 * e['weight'] / E) for e in g.es]
ID = max(g.vs.indegree()) or 1
g.vs['label_size'] = [max(12, 36 * v.indegree() / ID) for v in g.vs]
for v in g.vs:
if v['type'] == 'mention':
v['label_color'] = 'blue'
for v in g.vs:
filename = '%s/%s%s' % (FDAT_DIR, v['user_id'], FDAT_EXT)
if not os.path.isfile(filename):
v['color'] = (0, 0, 0)
filename = '_'.join(args.screen_name) + \
('_s' if args.strong else '_w') + \
('_t' if args.transparent else '') + \
'.' + args.format
igraph.plot(g,
filename,
layout=g.layout(args.layout),
bbox=(width, height),
margin=50)
def make_navigationoptions(args):
if not args.prefix:
args.prefix = os.path.splitext(os.path.basename(args.dataset))[0]
dataset = np.loadtxt(args.dataset)
graph = ig.load(args.graph)
vd = cPickle.load(open(args.clustering, 'rb'))
cl = vd.as_clustering()
index = FLANN()
index.load_index(args.index, dataset)
# Find adjacent communities
k = args.nn
knn, dists = index.nn_index(dataset, k+1)
knn = knn[:, 1:]
dists = dists[:, 1:]
membership = np.array(cl.membership)
adjacencies = {}
for community in cl:
for node in community:
mode_label = mode(membership[knn[node]])
mode_freq = mode_label[1][0]
mode_label = mode_label[0][0]
if mode_label != membership[node] and sum(graph.are_connected(node, nid) for nid in knn[node]) > (k/2):
adjacencies.setdefault(membership[node], []).append(int(mode_label))
# Make options
cPickle.dump([options.NavigationOption(source, target, membership, args.index, args.dataset, args.nn)
for source, targets in adjacencies.iteritems()
for target in set(targets)], open(args.prefix + '-navigation-options.pl', 'wb'))
def __prepare__(self):
"""
"""
#convert gml to graphml
G = igraph.load( os.path.join(self.raw_data_path, "football.gml"))
#must delete the id attribute since graphml uses it as a reserved attribute and gml does not
del G.vs['id']
G.write_graphml(self.graph_path)
def __prepare__(self):
"""
"""
#convert gml to graphml
G = igraph.load(os.path.join(self.raw_data_path, "football.gml"))
#must delete the id attribute since graphml uses it as a reserved attribute and gml does not
del G.vs['id']
G.write_graphml(self.graph_path)
def main(argv=None):
if argv==None:
argv = sys.argv
if len(argv) < 5 or len(argv) > 6:
print "Usage: %s songGraph.pickle exclusionList NumNoms src [commStruct]"% argv[0]
return
global GIVE
exclusionRaw = open(argv[2]).readlines()
exclusionList = []
for edge in exclusionRaw:
exclusionList.append(edge.strip())
numNoms = int(argv[3])
if len(argv) == 6:
# songReport = cPickle.load(open(argv[5]))
# commStruct = songReport.vcG
commStruct = cPickle.load(open(argv[5]))
g = commStruct.graph.copy()
else:
commStruct = None
g = igraph.load(argv[1])
if argv[4] == 'None':
src = None
paths = None
else:
src = argv[4]
omras2DB = pg.connect(dbname=DBNAME, host=HOST, user=USER)
try:
currentSession = omras2DB.query("SELECT session_no FROM sessions WHERE is_current = true").getresult()[0][0]
nomSession = int(currentSession)
print "%s :: Nominations will be inserted into session %i"%(argv[0], nomSession)
except IndexError:
print "%s :: no current session. turn on the radio dameon and try again."%argv[0]
return
print "%s:: removed %i edges from graph."%(argv[0], len(exclusionList))
g = g.delete_vertices(g.vs.select(hashkey_in=exclusionList))
toDelete = g.vs.select(duration=None)
print "%s:: removed %i vertices from graph due to unknown duration."%(argv[0], len(toDelete))
g = g.delete_vertices(toDelete)
toDelete = g.vs.select(hashkey=None)
print "%s:: removed %i vertices from graph due to unknown hashkey."%(argv[0], len(toDelete))
g = g.delete_vertices(toDelete)
deadEnds = g.vs.select(_outdegree=0)
try:
src = g.vs.select(hashkey= src)[0].index
except Exception, err:
print "%s :: ERROR :: trouble dereferencing src uid, nominees may be incorrectly filtered. \n\tmsg:%s\n\tproceeding w/o src"%(argv[0], str(err))
src = None
def new_top_k(net, k):
print('method called')
top_k_net = nx.Graph()
top_k = []
#loops through edges in network
for e in net.edges:
#defines end points for edges
source, target = e
#sorts edges for both end points
sorted_source = sorted(net.edges(source, data=True),
key=lambda t: t[2].get('cosine', 1),
reverse=True)
sorted_target = sorted(net.edges(target, data=True),
key=lambda t: t[2].get('cosine', 1),
reverse=True)
#checks if number of edges is within K already
if k < (len(sorted_source)):
top_source = sorted_source[:k]
else:
top_source = sorted_target
if k < (len(sorted_target)):
top_target = sorted_target[:k]
else:
top_target = sorted_target
#adds edge to top_k_net
for s in top_source:
s_source, s_target, s_cos = s
for t in top_target:
t_source, t_target, t_cos = t
if t_source == s_target and s_source == t_target:
top_k_net.add_edge(str(s_source),
str(s_target),
weight=s_cos)
#adds nodes
for n in nx.nodes(net):
top_k_net.add_node(n)
#export
nx.write_graphml(top_k_net, "graph.graphml")
nx.write_gml(top_k_net, "graph.gml")
graph = igraph.load("graph.gml")
layout = graph.layout("fr")
igraph.Graph.write_svg(graph,
fname="static/graph_image.svg",
layout=layout,
width=4000,
height=4000)
return top_k_net
def main():
import optparse
p = optparse.OptionParser(usage="usage: %prog [-s SOURCE] [-h] file")
p.add_option('-s', '--source', metavar='SOURCE', dest='source',
help='Specify a graph to use as source for attributes '+ \
'(this will disable API calls)')
opts, files = p.parse_args()
if not files:
p.error("Give me a file, please ;-)")
fn = files[0]
lang, date, type_ = explode_dump_filename(fn)
groups = ('bot', 'sysop', 'bureaucrat', 'checkuser', 'steward', 'import',
'transwiki', 'uploader', 'ipblock-exempt', 'oversight',
'founder', 'rollbacker', 'accountcreator', 'autoreviewer',
'abusefilter')
g = ig.load(fn)
if opts.source:
sourceg = ig.load(opts.source)
for destv in g.vs:
try:
sourcev = sourceg.vs.select(username=destv['username'])[0]
except IndexError:
print destv['username'], 'not found in source'
for group in groups:
destv[group] = None
continue
for group in groups:
destv[group] = sourcev[group]
else:
for group in groups:
addGroupAttribute(g, lang, group)
print 'BLOCKED ACCOUNTS'
addBlockedAttribute(g, lang)
print 'ANONYMOUS USERS'
g.vs['anonymous'] = map(isip, g.vs['username'])
g.write("%swiki-%s%s_rich.pickle" % (lang, date, type_), format="pickle")
def main():
import optparse
p = optparse.OptionParser(usage="usage: %prog [-s SOURCE] [-h] file")
p.add_option('-s', '--source', metavar='SOURCE', dest='source',
help='Specify a graph to use as source for attributes '+ \
'(this will disable API calls)')
opts, files = p.parse_args()
if not files:
p.error("Give me a file, please ;-)")
fn = files[0]
lang, date, type_ = explode_dump_filename(fn)
groups = ('bot', 'sysop', 'bureaucrat', 'checkuser', 'steward', 'import',
'transwiki', 'uploader', 'ipblock-exempt', 'oversight',
'founder', 'rollbacker', 'accountcreator', 'autoreviewer',
'abusefilter')
g = ig.load(fn)
if opts.source:
sourceg = ig.load(opts.source)
for destv in g.vs:
try:
sourcev = sourceg.vs.select(username=destv['username'])[0]
except IndexError:
print destv['username'], 'not found in source'
for group in groups:
destv[group] = None
continue
for group in groups:
destv[group] = sourcev[group]
else:
for group in groups:
addGroupAttribute(g, lang, group)
print 'BLOCKED ACCOUNTS'
addBlockedAttribute(g, lang)
print 'ANONYMOUS USERS'
g.vs['anonymous'] = map(isip, g.vs['username'])
g.write("%swiki-%s%s_rich.pickle" % (lang, date, type_), format="pickle")
def main():
facebook_graph = igraph.load('../data/fb_caltech_small_edgelist.txt', )
final_communities = sac1(facebook_graph)
file = open('communities.txt', 'w+')
for c in final_communities:
community = map(lambda x: str(x), c)
file.write(", ".join(community) + "\n")
file.close()
def loadFromIgraph(self, filename, gformat="graphml"):
"""
Load a sparse matrix from igraph as a numpy pickle
Positional arguments:
====================
filename - the file name/path to where you want to save the graph
gformat - the format which you want to use to save the graph. Choices:
"""
self.spcscmat = igraph.load(filename, format=gformat)
def loadFromIgraph(self, filename, gformat="graphml"):
"""
Load a sparse matrix from igraph as a numpy pickle
Positional arguments:
====================
filename - the file name/path to where you want to save the graph
gformat - the format which you want to use to save the graph. Choices:
"""
self.graph = igraph.load(filename, format=gformat)
def main():
facebook_graph = igraph.load('./data/fb_caltech_small_edgelist.txt')
final_communities = sac1(facebook_graph)
file = open('communities.txt', 'w+')
for c in final_communities:
community = map(lambda x: str(x), c)
file.write(", ".join(community) + "\n")
file.close()
def get_graph():
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = os.path.join(data_dir, "out_1.graphml")
if not os.path.exists(graph_path):
__download__(data_dir)
__prepare__(data_dir)
return igraph.load(graph_path)
def max_net(net, max_size):
#creates new list of edges to form filtered network
new_component_total = []
#identifies component and sorts edges by weight
for node in nx.nodes(net):
component = nx.node_connected_component(net, node)
#list of all edges in a component
component_edges = []
#adds to list of component edges
for n in component:
edges = sorted(net.edges(n, data=True),
key=lambda t: t[2].get('weight', 1),
reverse=True)
component_edges.extend(edges)
#if component size is > max_size, edges are removed until it is <= max_size and a new component is generated
if (len(component) > max_size):
new_component = make_new_component(
remove_last_edge(net, component_edges, node, max_size),
max_size)
new_component_total.extend(new_component)
#if a component is small enough, new_component total is automatically extended
else:
new_component_total.extend(edges)
#removes duplicates from edge list
component_final = []
for edge in new_component_total:
if edge not in component_final:
component_final.append(edge)
#creates a new network with filtered components
max_size_component_net = nx.Graph(component_final)
for n in nx.nodes(net):
max_size_component_net.add_node(n)
nx.write_graphml(max_size_component_net, "graph.graphml")
nx.write_gml(max_size_component_net, "graph.gml")
graph = igraph.load("graph.gml")
layout = graph.layout("fr")
igraph.Graph.write_svg(graph,
fname="static/graph_image.svg",
layout=layout,
width=4000,
height=4000)
return max_size_component_net
def load_trials(dir, rho):
files = os.listdir(dir)
ext = '.gml'
rword = 'rho=' + str(rho) + '_'
trialint = lambda s: int(s.strip(ext).split('_')[-1].split('=')[-1])
files = list(filter(lambda s: s.endswith(ext) and (rword in s), files))
files.sort(key=trialint)
gs = []
for fn in files:
gs.append(ig.load(dir + '/' + fn))
return gs
def setUp(self):
self.G=igraph.load("karate.gml")
membership=[
[0,1,2,3,7,11,12,13,17,19,21],
[4,5,6,10,16],
[8,9,14,15,18,20,22,23,24,25,26,27,28,29,30,31,32,33]]
cover=igraph.VertexCover(self.G, membership)
metrics=VertexCoverMetric.run_analysis(cover, weights=None)
metrics.report()
self.comm_metrics=metrics.comm_metrics
def get_graph():
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = os.path.join(data_dir, "out_1.graphml")
if not os.path.exists(graph_path):
__download__(data_dir)
__prepare__(data_dir)
return igraph.load(graph_path)
def main():
import igraph
if len(argv) == 2:
g = igraph.load(argv[1])
print g.summary()
c_im = g.community_infomap()
s_im = igraph_surprise(g, c_im)
print c_im.summary()
print "Infomap: Surprise = %s, Modularity = %s" % (s_im, c_im.q)
print
c_fg = g.as_undirected().community_fastgreedy().as_clustering()
s_fg = igraph_surprise(g, c_fg)
print c_fg.summary()
print "Fast Greedy: Surprise = %s, Modularity = %s" % (s_fg, c_fg.q)
print
if not g.is_directed():
c_mls = g.community_multilevel(return_levels=True)
else:
c_mls = g.as_undirected().community_multilevel(return_levels=True)
for i, c_ml in enumerate(c_mls):
s_ml = igraph_surprise(g, c_ml)
print c_ml.summary()
print "Multi-Level at Level %s: Surprise = %s, Modularity = %s" % (i, s_ml, c_ml.q)
print
elif len(argv) == 3:
g = igraph.Graph.Read_Edgelist(argv[1])
# load partition data into dict with node id as index
tmp = dict([ map(int,l.split()) for l in open(argv[2]).readlines() ])
# re-map to 0 based partition id's
zmap = dict((v, k) for k,v in enumerate(set(tmp.values())))
p = dict([(k, zmap[v]) for k,v in tmp.iteritems() ])
# load partition data into clustering object
# to re-use igraph functions
vc = igraph.VertexClustering(g, [ p[v.index] for v in g.vs ])
print g.summary()
print vc.summary()
# get S
s = igraph_surprise(g, vc)
print "Surprise = %s, Modularity = %s" % (s, vc.q)
else:
print "Usage: %s graph-file" % argv[0]
print " or %s network-edge-list partition-file" % argv[0]
def get_graph():
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = "nba.graphml"
full_path = os.path.join(data_dir, graph_path)
if not os.path.exists(data_dir):
__download__(data_dir)
if not os.path.exists(full_path):
__prepare__(data_dir, graph_path)
return igraph.load(full_path)
def get_graph():
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = "nba.graphml"
full_path = os.path.join(data_dir, graph_path)
if not os.path.exists(data_dir):
__download__(data_dir)
if not os.path.exists(full_path):
__prepare__(data_dir, graph_path)
return igraph.load(full_path)
def setUp(self):
self.G = igraph.load("karate.gml")
membership = [[0, 1, 2, 3, 7, 11, 12, 13, 17, 19, 21],
[4, 5, 6, 10, 16],
[
8, 9, 14, 15, 18, 20, 22, 23, 24, 25, 26, 27, 28, 29,
30, 31, 32, 33
]]
cover = igraph.VertexCover(self.G, membership)
metrics = VertexCoverMetric.run_analysis(cover, weights=None)
metrics.report()
self.comm_metrics = metrics.comm_metrics
def get_graph(self):
'''
Returns the graph loaded in memory
'''
if not os.path.exists(self.raw_data_path):
os.mkdir(self.raw_data_path)
self.__download__()
if not os.path.exists(self.graph_path):
self.__prepare__()
return igraph.load(self.graph_path)
def get_graph(self):
'''
Returns the graph loaded in memory
'''
if not os.path.exists(self.raw_data_path):
os.mkdir(self.raw_data_path)
self.__download__()
if not os.path.exists(self.graph_path):
self.__prepare__()
return igraph.load(self.graph_path)
def get_graph():
"""
Downloads and prepares the graph from DOWNLOAD_URL
"""
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = os.path.join(data_dir, GRAPH_NAME + GRAPH_TYPE)
if not os.path.exists(graph_path):
__download__(data_dir)
__prepare__(data_dir)
G = igraph.load(graph_path)
multigraph_to_weights(G)
return G
def get_graph():
"""
Downloads and prepares the network science collaboration graph
"""
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = os.path.join(data_dir, GRAPH_NAME + GRAPH_TYPE)
if not os.path.exists(graph_path):
__download__(data_dir)
__prepare__(data_dir)
else:
print(graph_path, "already exists. Using old file.")
return igraph.load(graph_path)
def get_graph():
"""
Downloads and prepares the network science collaboration graph
"""
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = os.path.join(data_dir, GRAPH_NAME + GRAPH_TYPE)
if not os.path.exists(graph_path):
__download__(data_dir)
__prepare__(data_dir)
else:
print(graph_path, "already exists. Using old file.")
return igraph.load(graph_path)
def __prepare__(self):
data = os.path.join(self.raw_data_path, DATA_DIR, "malaria.edgelist")
mod_data = os.path.join(self.raw_data_path, DATA_DIR, "mod_malaria.edgelist")
#we just need to remove the third column which has 1's in it
#so igraph can read it as an edgelist
with open(data, 'r') as f:
with open(mod_data, 'w') as new:
for line in f:
new.write(line[:-2] + '\n')
G = igraph.load(mod_data)
G.write_graphml(self.graph_path)
def get_graph():
"""
TEMPLATE COMMENT: Downloads and prepares a graph
"""
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = os.path.join(data_dir, GRAPH_NAME + GRAPH_TYPE)
if not os.path.exists(graph_path):
__download__(data_dir)
__prepare__(data_dir)
else:
print(graph_path, "already exists. Using old file.")
return igraph.load(graph_path)
def get_graph():
"""
TEMPLATE COMMENT: Downloads and prepares a graph
"""
data_dir = os.path.join(os.path.dirname(__file__), "data")
graph_path = os.path.join(data_dir, GRAPH_NAME + GRAPH_TYPE)
if not os.path.exists(graph_path):
__download__(data_dir)
__prepare__(data_dir)
else:
print(graph_path, "already exists. Using old file.")
return igraph.load(graph_path)
def main():
parser = argparse.ArgumentParser(
description=('Generate test graphs'),
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('graph_out', metavar='graph-out')
parser.add_argument('trace_out', metavar='trace-out')
parser.add_argument('--progressbar', action='store_true')
parser.add_argument('--verbose', '-v', action='count', default=0)
parser.add_argument('--node-count',
'-nc',
type=int,
dest='node_count',
default=100)
parser.add_argument('--network-type',
choices=[x for x in graphs_map.iterkeys()],
default='star')
parser.add_argument('--trace-count', type=int, default=50)
arguments = parser.parse_args()
arguments.verbose = min(len(helpers.LEVELS), arguments.verbose)
logging.getLogger('compnet').setLevel(helpers.LEVELS[arguments.verbose])
show_progress = arguments.progressbar
# g = graphs_map[arguments.network_type](arguments.node_count)
# g = igraph.Graph.Barabasi(900, 9)
# for n in g.vs:
# n['closeness'] = g.closeness(n)
# n['name'] = 'V%d' % n.index
# g.save(arguments.graph_out)
g = igraph.load(arguments.graph_out)
pairs = [
random.sample(xrange(0, g.vcount()), 2)
for x in xrange(0, arguments.trace_count)
]
pairs = [[g.vs[x[0]]['name'], g.vs[x[1]]['name']] for x in pairs]
traces = []
for p in pairs:
trace = random.choice(g.get_all_shortest_paths(p[0], p[1]))
trace = [g.vs[x]['name'] for x in trace]
traces.append(trace)
helpers.save_to_json(arguments.trace_out, traces)
def main():
"""Main function"""
if not os.path.exists(NETWORK_FILE):
print "Input network file is not present!"
sys.exit(1)
graph = ig.load(NETWORK_FILE)
graph['name'] = 'meneame'
general_analysis(graph)
community_analysis(graph)
save_degree_distribution(graph, IMAGES_FOLDER)
save_weights_distribution(graph, IMAGES_FOLDER)
def setUp(self):
self.G=igraph.load("karate.gml")
membership=[
[0,1,2,3,7,11,12,13,17,19,21],
[4,5,6,10,16],
[8,9,14,15,18,20,22,23,24,25,26,27,28,29,30,31,32,33]]
weights=[5,7,4,5,8,7,2,1,1,6,7,4,9,6,8,2,2,1,2,5,6,5,7,7,3,4,4,6,7,7,5,7,4,8,5,4,5,3,1,6,4,3,3,3,1,6,2,7,8,8,1,7,5,7,5,4,7,3,7,5,8,9,4,2,8,8,6,3,6,6,8,5,6,7,5,7,7,7]
self.G.es['weight'] = weights
self.cover=igraph.VertexCover(self.G, membership)
self.comm_metrics = None
if False:
vcm = VertexCoverMetric()
metrics=vcm.run_analysis(self.cover, weights = None)
self.comm_metrics = metrics.comm_metrics
def get(self, id):
"""Retrieves the dataset with the given ID from Nexus.
Dataset IDs are formatted as follows: the name of a dataset on its own
means that a single network should be returned if the dataset contains
a single network, or multiple networks should be returned if the dataset
contains multiple networks. When the name is followed by a dot and a
network ID, only a single network will be returned: the one that has the
given network ID. When the name is followed by a dot and a star, a
dictionary mapping network IDs to networks will be returned even if the
original dataset contains a single network only.
E.g., getting C{"karate"} would return a single network since the
Zachary karate club dataset contains one network only. Getting
C{"karate.*"} on the other hand would return a dictionary with one
entry that contains the Zachary karate club network.
@param id: the ID of the dataset to retrieve.
@return: an instance of L{Graph} (if a single graph has to be returned)
or a dictionary mapping network IDs to instances of L{Graph}.
"""
from igraph import load
dataset_id, network_id = self._parse_dataset_id(id)
params = dict(format="Python-igraph", id=dataset_id)
response = self._get_response("/api/dataset", params, compressed=True)
response = self._ensure_uncompressed(response)
result = load(response, format="pickle")
if network_id is None:
# If result contains a single network only, return that network.
# Otherwise return the whole dictionary
if not isinstance(result, dict):
return result
if len(result) == 1:
return result[result.keys()[0]]
return result
if network_id == "*":
# Return a dict no matter what
if not isinstance(result, dict):
result = dict(dataset_id=result)
return result
return result[network_id]
def get(self, id):
"""Retrieves the dataset with the given ID from Nexus.
Dataset IDs are formatted as follows: the name of a dataset on its own
means that a single network should be returned if the dataset contains
a single network, or multiple networks should be returned if the dataset
contains multiple networks. When the name is followed by a dot and a
network ID, only a single network will be returned: the one that has the
given network ID. When the name is followed by a dot and a star, a
dictionary mapping network IDs to networks will be returned even if the
original dataset contains a single network only.
E.g., getting C{"karate"} would return a single network since the
Zachary karate club dataset contains one network only. Getting
C{"karate.*"} on the other hand would return a dictionary with one
entry that contains the Zachary karate club network.
@param id: the ID of the dataset to retrieve.
@return: an instance of L{Graph} (if a single graph has to be returned)
or a dictionary mapping network IDs to instances of L{Graph}.
"""
from igraph import load
dataset_id, network_id = self._parse_dataset_id(id)
params = dict(format="Python-igraph", id=dataset_id)
response = self._get_response("/api/dataset", params, compressed=True)
response = self._ensure_uncompressed(response)
result = load(response, format="pickle")
if network_id is None:
# If result contains a single network only, return that network.
# Otherwise return the whole dictionary
if not isinstance(result, dict):
return result
if len(result) == 1:
return result[list(result.keys())[0]]
return result
if network_id == "*":
# Return a dict no matter what
if not isinstance(result, dict):
result = dict(dataset_id=result)
return result
return result[network_id]
def main(f_path):
global centrality_calculation_by_igraph, centrality_calculation_by_networkx
G = igraph.load(f_path)
centrality_calculation_by_igraph = c.centrality_calculation_by_igraph(G)
G = nx.DiGraph(G.get_edgelist())
centrality_calculation_by_networkx = c.centrality_calculation_by_networkx(
G)
# cluster_coefficient_calculation = c.cluster_coefficient_calculation(G)
return centrality_calculation_by_igraph,
centrality_calculation_by_networkx
def make_navigationoptions(args):
if not args.prefix:
args.prefix = os.path.splitext(os.path.basename(args.dataset))[0]
dataset = np.loadtxt(args.dataset)
graph = ig.load(args.graph)
vd = cPickle.load(open(args.clustering, 'rb'))
cl = vd.as_clustering()
index = FLANN()
index.load_index(args.index, dataset)
# Find adjacent communities
k = args.nn
knn, dists = index.nn_index(dataset, k + 1)
knn = knn[:, 1:]
dists = dists[:, 1:]
membership = np.array(cl.membership)
adjacencies = {}
for community in cl:
for node in community:
mode_label = mode(membership[knn[node]])
mode_freq = mode_label[1][0]
mode_label = mode_label[0][0]
if mode_label != membership[node] and sum(
graph.are_connected(node, nid)
for nid in knn[node]) > (k / 2):
adjacencies.setdefault(membership[node],
[]).append(int(mode_label))
# Make options
cPickle.dump([
options.NavigationOption(source, target, membership, args.index,
args.dataset, args.nn)
for source, targets in adjacencies.iteritems()
for target in set(targets)
], open(args.prefix + '-navigation-options.pl', 'wb'))
def igraph_to_networkx(netx):
'''returns an igraph instance,given all edges, vertices, and appropriate attributes'''
if isinstance(netx,str):
iG=igraph.load(netx)
else:
iG=netx
netG=networkx.Graph()
int_to_id={}
for i in range(len(iG.vs)):
attr=iG.vs[i].attributes()
int_to_id[i]=attr["id"]
netG.add_node(attr["id"],attr_dict=attr)
del attr["id"]
#print(int_to_id)
for i in range(len(iG.es)):
m=int_to_id[iG.es[i].source]
n=int_to_id[iG.es[i].target]
netG.add_edge(m,n,iG.es[i].attributes())
return netG
def result():
print(request.form)
username=request.form.get("username")
password=request.form.get("password")
analysis_id=int(request.form.get("analysis_id"))
top_k=request.form.get("top_k")
max_component_size=request.form.get("max_component_size")
filter = request.form.get("filters")
spectra_list, matches = sim.compare(pimport.import_pimp(username, password, analysis_id))
#generates network by passing spectra_list(nodes) and matches(edges)
network = networking.networking(spectra_list, matches)
#applies filters
if filter is not None:
if top_k != "":
network = networking.new_top_k(network, (int(top_k)))
if max_component_size != "":
network = networking.max_net(network, (int(max_component_size)))
#networking.graphml(network, "graph.graphml")
#export
nx.write_graphml(network, "uploads/graph.graphml")
nx.write_gml(network, "graph.gml")
graph = igraph.load("graph.gml")
layout = graph.layout("kk")
igraph.Graph.write_svg(graph,fname="static/graph_image.svg", layout = layout, width = 4000, height = 4000)
graph_file = "graph_image.svg"
return render_template("result.html", graph_image = graph_file)
def __prepare__(self):
vertex_filename = os.path.join(self.raw_data_path, VERTEX_DATA_FILE)
edge_filename = os.path.join(self.raw_data_path, EDGE_DATA_FILE)
g = igraph.load(edge_filename)
vertex_file = open(vertex_filename)
reader = csv.DictReader(vertex_file)
for case in reader:
caseid = int(case['caseid'])
v = g.vs[caseid]
v['caseid'] = case['caseid']
v['usid'] = case['usid']
v['parties'] = case['parties']
v['year'] = case['year']
vertex_file.close()
# Case IDs are 1-indexed, so we delete the 0th vertex as it was extraneous.
g.delete_vertices([0])
g.write_graphml(self.graph_path)
def example():
"""
Performs the example outlined in the README. Draws the graph of one dataset.
"""
g = ig.load("sp_data_school_day_1_g.graphml") # whichever file you would like
# Assigning colors to genders for plotting
colorDict = {"M": "blue", "F": "pink", "Unknown": "black"}
for vertex in g.vs:
# each vertex is labeled as its classname and colored as its gender.
vertex["label"] = vertex["classname"]
vertex["color"] = colorDict[vertex["gender"]]
layout = g.layout("fr") # Fruchterman-Reingold layout
# If Cairo is improperly installed, raises TypeError: plotting not available
try:
ig.plot(g, layout=layout)
except TypeError as e:
print "It looks like Cairo isn't properly installed. Refer to the wiki."
exit(1)
def main():
### Read initial graph
#g = igraph.load('data_repo/Imputed_data_False9_0509.graphmlz') ### This file contains the weekday 9am link level travel time for SF, imputed data collected from a month worth of Google Directions API
g = igraph.load('data_repo/London_Directed/London_0621.graphmlz')
print(g.summary())
mode = 'ALL' ### 'IN', 'OUT', 'ALL'
degree_dist = g.degree_distribution(mode=mode)
loc = []
val = []
for bn in degree_dist.bins():
loc.append(bn[0])
val.append(bn[2])
print(loc, val)
plt.bar(loc, val, width=1, color='w', edgecolor='black', log=True)
for i, v in enumerate(val):
plt.text(i - 0.25, v, str(v), color='blue', fontweight='bold')
plt.xlim([-1, (int(max(loc)) + 1)])
plt.title('Degree distribution of mode {}'.format(mode))
plt.xlabel('Degree')
plt.ylabel('Vertices count')
plt.show()
def label_igraph_network(network_path):
g = i.load(network_path)
g = g.as_undirected()
print 'Get giant component'
g = g.clusters(mode='weak').giant()
simple = g.is_simple()
print 'IS SIMPLE? %s' % simple
if not simple:
g.simplify(multiple=True, loops=True, combine_edges='ignore')
try:
for vs in g.vs:
vs['name'] = 'LBL%s' % vs['name']
except KeyError:
for vs in g.vs:
vs['name'] = 'LBL%s' % vs.index
edge_list = helpers.degree_labeling_network(g, 1.1)
vs = list(set([y for t in [(x[0], x[1]) for x in edge_list] for y in t]))
labeled_g = helpers.load_as_inferred_links_nofile(vs, edge_list)
return labeled_g
def __prepare__(self):
vertex_filename = os.path.join(self.raw_data_path, VERTEX_DATA_FILE)
edge_filename = os.path.join(self.raw_data_path, EDGE_DATA_FILE)
g = igraph.load(edge_filename)
vertex_file = open(vertex_filename)
reader = csv.DictReader(vertex_file)
for case in reader:
caseid = int(case['caseid'])
v = g.vs[caseid]
v['caseid'] = case['caseid']
v['usid'] = case['usid']
v['parties'] = case['parties']
v['year'] = case['year']
vertex_file.close()
# Case IDs are 1-indexed, so we delete the 0th vertex as it was extraneous.
g.delete_vertices([0])
g.write_graphml(self.graph_path)
def process_file(self, filename):
"""Loads a graph from the given file, runs the clustering
algorithm on it and prints the clusters to the standard
output."""
self.log.info("Processing %s..." % filename)
graph = load(filename, format=self.options.format)
# If the graph has weights and we want to ignore them, delete them
if self.options.no_weights and "weight" in graph.edge_attributes():
del graph.es["weight"]
# If the graph is directed, we have to make it undirecteed
if graph.is_directed():
graph.to_undirected(combine_edges="sum")
self.log.warning("Converted directed graph to undirected.")
# Set up the "name" attribute properly
if "label" in graph.vertex_attributes():
graph.vs["name"] = graph.vs["label"]
del graph.vs["label"]
elif "name" not in graph.vertex_attributes():
graph.vs["name"] = [str(i) for i in xrange(graph.vcount())]
# Run the algorithm, get the result generator
self.log.info("Calculating clusters, please wait...")
algorithm = HLC(graph, self.options.min_size)
results = algorithm.run(self.options.threshold)
# Print the optimal threshold if we determined it automatically
if self.options.threshold is None:
self.log.info("Threshold = %.6f" % algorithm.last_threshold)
self.log.info("D = %.6f" % algorithm.last_partition_density)
# Print the results
for community in results:
print "\t".join(graph.vs[community]["name"])
def main(argv=None):
if argv==None:
argv = sys.argv
if len(argv) < 2 or len(argv) > 3:
print "Usage: %s songGraph.pkl [vertexDendrogram.pickle]"% argv[0]
return
if len(argv) == 3:
vertClustPkl = argv[2]
else:
vertClustPkl = None
s = shout.Shout()
print "Using libshout version %s" % shout.version()
s.host = 'doc.gold.ac.uk'
s.port = 8000
s.user = '******'
s.password = '******'
s.mount = "/vRad.mp3"
s.name = "Steerable Optimized Self-Organizing Radio"
s.genre = "CompNets"
s.url = "http://radio.benfields.net"
s.format = 'mp3' # | 'vorbis'
#s.protocol = 'icy' | 'xaudiocast' | 'http'
# s.public = 0 | 1
# s.audio_info = { 'key': 'val', ... }
# (keys are shout.SHOUT_AI_BITRATE, shout.SHOUT_AI_SAMPLERATE,
# shout.SHOUT_AI_CHANNELS, shout.SHOUT_AI_QUALITY)
currentTracklist = []
try:
graphLocation = argv[1]
loadedGraph = igraph.load(graphLocation, format="pickle")
ms = copy.deepcopy(loadedGraph)
except Exception, err:
print "trouble loading the graph at {0} should point to the graph (as pickle). Sort it out and try again.".format(argv[1])
return
def __prepare__(self):
G = igraph.load(os.path.join(self.raw_data_path, "netscience.gml"))
del G.vs['id'] #graphml uses the id field, so we must remove it
G.write_graphml(self.graph_path)
def get_books(self):
g = ig.load("territories/data/polbooks.gml")
g.simplify(loops=False)
return g