def powerlaw_degree_sequence(n, a):
"""
Create a graph without self-loops or parallel edges; having power law degree
distribution with an exponent 'around' a.
Parameters
----------
n : int
Number of nodes in graph.
a : float
Ideal exponent.
Returns
-------
G : networkx.Graph
The constructed graph.
"""
dsq = nx.create_degree_sequence(n, nx.utils.powerlaw_sequence, exponent = a)
G = nx.Graph(nx.configuration_model(dsq))
G.remove_edges_from(G.selfloop_edges())
# Check for a disconnected graph just in case...
if not nx.is_connected(G):
emsg = "The generated power-law graph is not connected!"
logger.error(emsg)
raise NepidemiXBaseException(emsg)
return G
def configuration_model():
func = lambda n:nx.utils.powerlaw_sequence(n, 4.0)
z=nx.create_degree_sequence(100, func)
G=nx.configuration_model(z)
G=nx.Graph(G)
G.remove_edges_from(G.selfloop_edges())
G.name = 'power law'
return G
""" Random graph from given degree sequence. Draw degree histogram with matplotlib. """ __author__ = """Aric Hagberg ([email protected])""" try: import matplotlib.pyplot as plt import matplotlib except: raise import networkx as nx z=nx.create_degree_sequence(100,nx.utils.powerlaw_sequence,exponent=2.1) nx.is_valid_degree_sequence(z) print "Configuration model" G=nx.configuration_model(z) # configuration model degree_sequence=sorted(nx.degree(G).values(),reverse=True) # degree sequence #print "Degree sequence", degree_sequence dmax=max(degree_sequence) plt.loglog(degree_sequence,'b-',marker='o') plt.title("Degree rank plot") plt.ylabel("degree") plt.xlabel("rank") # draw graph in inset
if __name__ == '__main__':
import time
try:
import igraph
except ImportError:
# it would be nice to implement biconnected_components in NetworkX
# it is in my TODO list
raise Exception('Unable to import igraph (used only to test accuracy)')
Gnp = networkx.gnp_random_graph(100, 0.03)
Gba = networkx.barabasi_albert_graph(100, 2)
Gpc = networkx.powerlaw_cluster_graph(100, 2, 0.1)
constructor = [(20, 40, 0.8), (80, 140, 0.6)]
Gshell = networkx.random_shell_graph(constructor)
Gshell.name = 'Shell graph'
deg_seq = networkx.create_degree_sequence(100,
networkx.utils.powerlaw_sequence)
Gconf = networkx.Graph(networkx.configuration_model(deg_seq))
Gconf.remove_edges_from(Gconf.selfloop_edges())
Gconf.name = 'Conf model'
Gdeb_2m = networkx.read_adjlist('test_2m.adj') # file in ticket #589
Gdeb_2m.name = "Debian 2-mode"
Gdeb_1m = networkx.read_adjlist('test_1m.adj') # file in ticket #589
Gdeb_1m.name = "Debian 1-mode"
graph_list = [Gnp, Gba, Gpc, Gshell, Gconf, Gdeb_1m, Gdeb_2m]
for G in graph_list:
print("Testing with %s" % G.name)
print(networkx.info(G))
print('Running analysis ...')
print
print("New version using the complement graph in step 3")
start = time.time()
if __name__ == '__main__':
import time
try:
import igraph
except ImportError:
# it would be nice to implement biconnected_components in NetworkX
# it is in my TODO list
raise Exception('Unable to import igraph (used only to test accuracy)')
Gnp = networkx.gnp_random_graph(100, 0.03)
Gba = networkx.barabasi_albert_graph(100, 2)
Gpc = networkx.powerlaw_cluster_graph(100, 2, 0.1)
constructor=[(20,40,0.8),(80,140,0.6)]
Gshell = networkx.random_shell_graph(constructor)
Gshell.name = 'Shell graph'
deg_seq = networkx.create_degree_sequence(100, networkx.utils.powerlaw_sequence)
Gconf = networkx.Graph(networkx.configuration_model(deg_seq))
Gconf.remove_edges_from(Gconf.selfloop_edges())
Gconf.name = 'Conf model'
Gdeb_2m = networkx.read_adjlist('test_2m.adj') # file in ticket #589
Gdeb_2m.name = "Debian 2-mode"
Gdeb_1m = networkx.read_adjlist('test_1m.adj') # file in ticket #589
Gdeb_1m.name = "Debian 1-mode"
graph_list = [Gnp, Gba, Gpc, Gshell, Gconf, Gdeb_1m, Gdeb_2m]
for G in graph_list:
print("Testing with %s"%G.name)
print(networkx.info(G))
print('Running analysis ...')
print
print("New version using the complement graph in step 3")
start = time.time()
def main():
# Base parameters
num_runs=500
num_agents=150
# Set up directory structure for data storage
data_dir="ABM_data" # Directory for all ABM data outout
sub_dirs={"binom":"binomial","uni":"uniform","pref":"pref_attach","par":"pareto","pl":"power_law"} # All sub-directries
# Create directories
try:
os.mkdir(data_dir)
except(OSError):
pass
for d in sub_dirs.values():
try:
os.mkdir(data_dir+"/"+d)
except(OSError):
pass
###### SIMULATION RUNS ######
for r in xrange(num_runs):
### Group 1: Binomial random networks ###
# Create random GNP network, and get degree sequence
gnp=nx.generators.gnp_random_graph(num_agents,p=.5)
binom_ds=gnp.degree()
E_BINOM=SB.Environment(population=num_agents,degree_seq=binom_ds)
E_BINOM.get_data(data_dir+"/"+sub_dirs["binom"]+"/"+str(r)+"_"+sub_dirs["binom"]+".csv")
### Group 2: Uniform degree distirbution random networks ###
# Create unidform degree sequence
uni_ds=nx.create_degree_sequence(num_agents,nx.utils.uniform_sequence)
E_UNI=SB.Environment(population=num_agents,degree_seq=uni_ds)
E_UNI.get_data(data_dir+"/"+sub_dirs["uni"]+"/"+str(r)+"_"+sub_dirs["uni"]+".csv")
### Group 3: Prefential attachment by agent wealth parameter (class default) ###
E_PREF=SB.Environment(population=num_agents)
E_PREF.get_data(data_dir+"/"+sub_dirs["pref"]+"/"+str(r)+"_"+sub_dirs["pref"]+".csv")
### Group 4: Pareto degree distribution random networks ###
par_ds=nx.create_degree_sequence(num_agents,nx.utils.pareto_sequence)
E_PAR=SB.Environment(population=num_agents,degree_seq=par_ds)
E_PAR.get_data(data_dir+"/"+sub_dirs["par"]+"/"+str(r)+"_"+sub_dirs["par"]+".csv")
### Group 5: Power-law degree distribution random networks ###
pl_ds=nx.create_degree_sequence(num_agents,nx.utils.powerlaw_sequence)
E_PL=SB.Environment(population=num_agents,degree_seq=pl_ds)
E_PL.get_data(data_dir+"/"+sub_dirs["pl"]+"/"+str(r)+"_"+sub_dirs["pl"]+".csv")
# Update the stdout on progress. Dumb method, but straightforward
if r==num_runs*.25:
# 25% complete
print "Simulation "+str((float(r)/num_runs)*100)+"% complete"
else:
if r==num_runs*.5:
print "Simulation "+str((float(r)/num_runs)*100)+"% complete"
else:
if r==num_runs*.75:
print "Simulation "+str((float(r)/num_runs)*100)+"% complete"
print "SIMULATION COMPLETE"
print ""
# Create single CSV file from all saved files for each network type
for d in sub_dirs.values():
for r in xrange(num_runs):
dr=csv.DictReader(open(data_dir+"/"+d+"/"+str(r)+"_"+d+".csv","r"))
if r==0:
dw=csv.DictWriter(open(data_dir+"/"+d+"/FULL_"+d+".csv", "w"),fieldnames=dr.fieldnames)
header=dict(zip(dr.fieldnames,dr.fieldnames))
dw.writerow(header)
for row in dr:
dw.writerow(row)
### Finally, archive data ###
print "Archiving data"
# Zip data files into single file
makeArchive(dirEntries(data_dir,True),data_dir+".zip")
""" Random graph from given degree sequence. Draw degree histogram with matplotlib. """ __author__ = """Aric Hagberg ([email protected])""" try: import matplotlib.pyplot as plt import matplotlib except: raise import networkx as nx z=nx.create_degree_sequence(400,nx.utils.powerlaw_sequence,exponent=2.1) nx.is_valid_degree_sequence(z) print "Configuration model" G=nx.configuration_model(z) # configuration model degree_sequence=sorted(nx.degree(G),reverse=True) # degree sequence #print "Degree sequence", degree_sequence dmax=max(degree_sequence) plt.loglog(degree_sequence,'b-',marker='o') plt.title("Degree rank plot") plt.ylabel("degree") plt.xlabel("rank") # draw graph in inset