def small_worldness(self, G, N, matrix):
G_rand = nx.gnm_random_graph(G.number_of_nodes(), G.number_of_edges())
path_length = nx.average_shortest_path_length(G, weight='weight')
# G_rand path length
weight_matrix = copy(matrix)
if 'edge_threshold' in self.params:
edge_threshold = self.params['edge_threshold']
else:
edge_threshold = 0.
non_edges = np.where(weight_matrix <= edge_threshold)
weight_matrix[non_edges[0], non_edges[1]] = 0.
np.fill_diagonal(weight_matrix, 0)
triu_idx = np.triu_indices(N, 1)
rand_weights = weight_matrix[triu_idx[0], triu_idx[1]]
rand_weights = np.array([w for w in rand_weights if w])
np.random.shuffle(rand_weights)
# print len(rand_weights), len(G_rand.edges())
for count, e in enumerate(G_rand.edges()):
G_rand[e[0]][e[1]]['weight'] = rand_weights[count]
path_length_rand = nx.average_shortest_path_length(G_rand,
weight='weight')
# transitivity
transitivity = self.transitivity(G, N, matrix)
# G_rand transitivity
G_rand2 = nx.gnm_random_graph(G.number_of_nodes(), G.number_of_edges())
transitivity_rand = nx.transitivity(G_rand2)
self.prediction_dim = 0
return (transitivity / transitivity_rand) / (path_length /
path_length_rand)
def gen_single_graph(pos, attr, dist_key):
global group_counter
global config
global attribute
s_levels = range(1, 6)
u_levels = range(1, 6)
G = nx.gnm_random_graph(18, 25)
while not nx.is_connected(G):
G = nx.gnm_random_graph(18, 25)
for e in G.edges():
G[e[0]][e[1]]['strength'] = random.choice(s_levels)
G[e[0]][e[1]]['certainty'] = random.choice(u_levels)
counter = 0
attr_values = list(dist_template[dist_key])
random.shuffle(attr_values)
for e in G.edges():
G[e[0]][e[1]][attr] = attr_values[counter]
counter += 1
# write json formatted data
d = json_graph.node_link_data(G) # node-link format to serialize
# write json
output_name = 'task_cp_' + attribute + '_' + config + '_' + str(group_counter) + '_' + pos + '.json'
output = open(output_name, 'w')
json.dump(d, output, indent=2, separators=(', ', ':' ))
print('Wrote node-link JSON data to ' + output_name)
def gen_single_graph(pos, attr, dist_key):
global group_counter
global config
global attribute
s_levels = range(1, 6)
u_levels = range(1, 6)
G = nx.gnm_random_graph(18, 25)
while not nx.is_connected(G):
G = nx.gnm_random_graph(18, 25)
for e in G.edges():
G[e[0]][e[1]]['strength'] = random.choice(s_levels)
G[e[0]][e[1]]['certainty'] = random.choice(u_levels)
counter = 0
attr_values = list(dist_template[dist_key])
random.shuffle(attr_values)
for e in G.edges():
G[e[0]][e[1]][attr] = attr_values[counter]
counter += 1
# write json formatted data
d = json_graph.node_link_data(G) # node-link format to serialize
# write json
output_name = 'task_cp_' + attribute + '_' + config + '_' + str(
group_counter) + '_' + pos + '.json'
output = open(output_name, 'w')
json.dump(d, output, indent=2, separators=(', ', ':'))
print('Wrote node-link JSON data to ' + output_name)
def setup(
network='random', # ['complete', 'random', 'Watts-Strogatz', 'connected caveman', 'Barabasi-Albert']
n_agents=40, # number of agents
deg=4, # number of connections for each agent
n_beliefs=25, # number of knowledge graph elements each agent has
n_concepts=30):
"""
Generates the initial conditions of the simulation
Returns
-------
g: networkx graph
primary graph represents the semantic network,
each individual (node) has an attribute 'M' representing their semantic network
all_beliefs: an array of tuples, which represents all edges anywhere in the semantic
network network of any individual. Does not include edges that could be part of the
semantic network (because they are present in a complete graph with `n_concepts`,
but were not selected).
"""
np.random.seed()
connected = False
while not connected:
if network == 'complete':
g = nx.complete_graph(n=n_agents)
elif network == 'random':
g = nx.gnm_random_graph(n=n_agents, m=int(n_agents * deg / 2))
elif network == 'random regular':
g = nx.random_regular_graph(d=deg, n=n_agents)
elif network == 'Watts-Strogatz':
g = nx.connected_watts_strogatz_graph(n=n_agents, k=deg,
p=.02) # valid for even deg
elif network == 'connected caveman':
g = nx.connected_caveman_graph(l=int(n_agents / deg), k=deg + 1)
elif network == 'Barabasi-Albert':
g = nx.barabasi_albert_graph(n=n_agents, m=int(
deg /
2)) # approximates deg for large n_agents, when deg is even.
else:
raise ValueError('%s is not a valid network name' % network)
connected = nx.is_connected(g)
# give starting information
nx.set_node_attributes(
g,
name='M', # M for mind
values={i: nx.gnm_random_graph(n_concepts, n_beliefs)
for i in g})
beliefs = np.unique([
tuple(sorted(belief)) for agent in g
for belief in g.nodes[agent]['M'].edges()
],
axis=0)
return g, beliefs
def generate_random_graph(n_nodes,
density,
directed=False,
is_tree=False,
seed=None,
draw=False):
r"""Generate a random graph with the desired number of nodes and density.
Draw the graph if asked. Returns the graph as a networkx object.
Parameters:
----------
n_nodes: int,
number of nodes
density: float (\in [0,1]),
density of edges
directed: bool,
whether or not the graph is directed
is_tree: bool,
whether to generate a tree or not
seed: int,
the random seed to use
draw: bool,
whether to draw the graph or not
Returns:
-------
graph: networkx Digraph"""
# if we want to generate a directed graph, we generate two undirected graphs
# of the same size and assemble them in one single directed graph by saying the
# vertices of the first are in the direction left -> right and the inverse for the second
# Compute the number of edges corresponding to the given density
n_edges = int(density * n_nodes * (n_nodes - 1) / 2)
# Create the graph
seed_1 = None if seed is None else seed + 1
if is_tree:
graph_0 = nx.random_tree(n_nodes, seed=seed)
if directed:
graph_1 = nx.random_tree(n_nodes, seed=seed_1)
else:
graph_1 = graph_0
else:
graph_0 = nx.gnm_random_graph(n_nodes, n_edges, seed)
if directed:
graph_1 = nx.gnm_random_graph(n_nodes, n_edges, seed_1)
else:
graph_1 = graph_0
graph_2 = nx.DiGraph()
graph_2.add_nodes_from(graph_0.nodes())
graph_2.add_edges_from([(u, v) for (u, v) in graph_0.edges()])
graph_2.add_edges_from([(v, u) for (u, v) in graph_1.edges()])
if draw:
plot_graph(graph_2, color_type="fabulous")
return graph_2
def getGraph():
edges1 = random.randint(nodeNo, nodeNo + 10)
graph1 = nx.gnm_random_graph(nodeNo, edges1, False)
edges2 = random.randint(nodeNo, nodeNo + 10)
graph2 = nx.gnm_random_graph(nodeNo, edges2, False)
return graph1, graph2
def gen_input(n,m): from random import randint import networkx as nx G = nx.gnm_random_graph(n, m) # gen graph while not nx.is_connected(G): G = nx.gnm_random_graph(n, m) # gen graph R = [randint(0, n) for i in xrange(n)] # gen requirements S = [randint(0, 100) for i in xrange(n)] # gen requirements return G, R, S
def test_different_number_of_edges(self):
for i in range(10):
N = np.random.randint(2, 50)
E1 = np.random.randint(N * (N - 1) / 2 - 1)
E2 = np.random.randint(N * (N - 1) / 2 - 1)
if E1 == E2:
E1 -= 1
G = nx.gnm_random_graph(N, E1)
H = nx.gnm_random_graph(N, E2)
assert not nx.is_isomorphic(G, H)
assert not is_possibly_isomorphic(G, H, N)
def BinarySearchMethod( N, L ):
locN,locL=N, L
G=nx.gnm_random_graph(locN,locL)
delta=1
# Check if we are lucky...
if __getGCCNodeCount(G)==locN:
return G
finished=False
while not finished:
while __getGCCNodeCount(G)<locN:
G=nx.gnm_random_graph(locN+delta,locL)
delta=delta*2
if __getGCCNodeCount(G)>=locN:
delta=int(delta/2)
left,right=int(locN+delta/2),int(locN+delta)
while left!=right:
middle=left+(right-left)/2
G=nx.gnm_random_graph(int(middle),locL)
NC=__getGCCNodeCount(G)
if NC>locN:
right=middle
elif NC<locN:
left=middle
else:
left=right
Gccs=sorted(nx.connected_component_subgraphs(G), key=len, reverse=True)
G0=Gccs[0]
nodes=list(G0.nodes())
while G.number_of_edges()<locL:
a=random.choice(nodes)
b=random.choice(nodes)
G0.add_edge(a,b)
curN,curL,curCon=G0.number_of_nodes(),G0.number_of_edges(),nx.is_connected(G0)
if curN==locN and curL==locL and curCon==True:
finished=True
return G0
def randomize_graph(self, n, e):
"""
Creates 1000 random graphs with
networkx.gnm_random_graph(nodecount, edgecount),
and computes average_clustering_coefficient and
average_shortest_path_length, to compare with drama-graph.
Returns a tuple:
randavgpathl, randcluster = (float or "NaN", float or "NaN")
"""
randcluster = 0
randavgpathl = 0
# what is c, what is a, what is n, what is e?
# e=edges?, c=clustering_coefficient?, a=average_shortest_path_length?
c = 0
a = 0
if not self.randomization: # hack so that quartett poster works
self.randomization = 50
for i in tqdm(range(self.randomization), desc="Randomization",
mininterval=1):
R = nx.gnm_random_graph(n, e)
try:
randcluster += nx.average_clustering(R)
c += 1
except ZeroDivisionError:
pass
j = 0
while True:
j += 1
try:
R = nx.gnm_random_graph(n, e)
randavgpathl += nx.average_shortest_path_length(R)
a += 1
except:
pass
else:
break
if j > 50:
randavgpathl = "NaN"
break
try:
randcluster = randcluster / c
except:
randcluster = "NaN"
try:
randavgpathl = randavgpathl / a
except:
randavgpathl = "NaN"
return randavgpathl, randcluster
def MinDistanceGraphFinder(number_of_vertices,number_of_edges): if number_of_edges<number_of_vertices-1: print 'Too few edges. The graph will be disconnected. Exiting...' return -1 if number_of_edges>(number_of_vertices-1)*number_of_vertices/2.0: print 'Too many edges. Such a graph cannot exist. Exiting...' return -1 if number_of_edges>(number_of_vertices-2)*(number_of_vertices-1)/2.0: OutputGraph=nx.gnm_random_graph(number_of_vertices,number_of_edges) return OutputGraph G=nx.star_graph(number_of_vertices-1) VertexList=G.nodes() remaining_edges=number_of_edges-number_of_vertices+1 while remaining_edges>0: first_vertex=random.choice(VertexList) second_vertex=random.choice(VertexList) if first_vertex==second_vertex:continue if (first_vertex,second_vertex) in G.edges():continue if (second_vertex,first_vertex) in G.edges():continue G.add_edge(first_vertex,second_vertex) remaining_edges-=1 OutputGraph=G.copy() return OutputGraph
def measure_single(graph_size):
global args
time_results = []
memory_results = []
print('graph size={}:'.format(graph_size), end=' ')
for density in args.densities:
vertices = compute_vertices_count(graph_size, density)
edges = graph_size - vertices
nx_graph = nx.gnm_random_graph(vertices, edges, directed=True)
if nx_graph.number_of_nodes() + nx_graph.number_of_edges() != graph_size:
time_results.append(float('nan'))
memory_results.append(float('nan'))
continue
print('[D={} |V|={} |E|={}]'.format(density, nx_graph.number_of_nodes(), nx_graph.number_of_edges()), end=' ')
stats = get_stats(args.repetitions, args.multiplication, [args.algorithm], nx_graph, args.garbage_collection)
time_results.append(round(min(stats[args.algorithm]['time']), 5))
memory_results.append(stats[args.algorithm]['memory'] / float(2**20))
print()
for res in [(time_results, '_time.csv'), (memory_results, '_memory.csv')]:
with open(args.output + res[1], 'a') as csvfile:
writer = csv.writer(csvfile)
row = [graph_size] + res[0]
writer.writerow(row)
def measure_multi(graph_size):
global args
vertices = compute_vertices_count(graph_size, args.density)
edges = graph_size - vertices
nx_graph = nx.gnm_random_graph(vertices, edges, directed=True)
if nx_graph.number_of_nodes() + nx_graph.number_of_edges() != graph_size:
return
print('graph size={}: [D={} |V|={} |E|={}]'.format(graph_size, args.density,
nx_graph.number_of_nodes(), nx_graph.number_of_edges()))
stats = get_stats(args.repetitions, args.multiplication, args.algorithms, nx_graph, args.garbage_collection)
#extract minimal time from all measured times
for key in stats:
stats[key]['time'] = round(min(stats[key]['time']), 5)
#write time and memory values into files as CSV
for res_type in ['time', 'memory']:
with open(args.output + '_' + res_type + '.csv', 'a') as csvfile:
writer = csv.writer(csvfile)
row = [graph_size] + [stats[key][res_type] for key in sorted(stats)]
writer.writerow(row)
def generate_random_graph(self):
print 'performer.generate_random_graph:', self.NODE_COUNT, self.EDGE_COUNT
while True:
graph = gnm_random_graph(self.NODE_COUNT, self.EDGE_COUNT)
if self.constraints_satisfied(graph):
self.process_graph(graph)
return graph
def test_layouts():
G =nx.gnm_random_graph(10,15)
rand = [nx.random_layout(G)]
circ = [nx.circular_layout(G)]
#shell = [nx.shell_layout(G)] #same as circular layout...
spectral = [nx.spectral_layout(G)]
tripod = [tripod_layout(G)]
layouts = [rand,circ,spectral, tripod]
regimes = ["random","circular","spectral", "tripod"]
for layout in layouts:
layout.append(nx.spring_layout(G,2,layout[0]))
layout.append(iterate_swaps(G,layout[0]))
layout.append(nx.spring_layout(G,2,layout[2]))
layout.append(greedy_swapper(G,layout[0]))
# Now have list of lists... Find lengths of edgecrossings...
num_crossings = []
for layout in layouts:
for sublayout in layout:
num_crossings.append(count_crosses(G,sublayout))
names = []
for regime in regimes:
names.append(regime)
names.append(regime + "-spring")
names.append(regime + "-swap")
names.append(regime + "-swap-spr")
names.append(regime + "-greedy")
return G, layouts, names, num_crossings
def generate_graph(self):
n = self._num_of_nodes
m = self._num_of_edges
G = networkx.gnm_random_graph(n, m)
for u, v in G.edges():
G[u][v]["weight"] = 1
return G
def create_graph(self, params):
# case statements on type of graph
if self.type == 'small_world':
self.graph = nx.watts_strogatz_graph(
int(params[0]), int(params[1]), float(params[2]),
None if len(params) < 4 else int(params[3]))
cg.SWC += 1
self.idx = cg.SWC
self.path = 'smallWorldGraphs'
elif self.type == 'small_world_connected':
self.graph = nx.connected_watts_strogatz_graph(
int(params[0]), int(params[1]), float(params[2]),
100 if len(params) < 4 else int(params[3]),
None if len(params) < 5 else int(params[4]))
cg.SWCC += 1
self.idx = cg.SWCC
self.path = 'smallWorldConnectedGraphs'
elif self.type == 'small_world_newman':
self.graph = nx.newman_watts_strogatz_graph(
int(params[0]), int(params[1]), float(params[2]),
None if len(params) < 4 else int(params[3]))
cg.SWNC += 1
self.idx = cg.SWNC
self.path = 'smallWorldNewmanGraphs'
elif self.type == 'power_law':
self.graph = nx.powerlaw_cluster_graph(
int(params[0]), int(params[1]), float(params[2]),
None if len(params) < 4 else int(params[3]))
cg.PLC += 1
self.idx = cg.PLC
self.path = 'powerLawGraphs'
elif self.type == 'power_law_tree':
self.graph = nx.random_powerlaw_tree(
int(params[0]), 3 if len(params) < 2 else float(params[1]),
None if len(params) < 3 else int(params[2]),
100 if len(params) < 4 else int(params[3]))
cg.PLTC += 1
self.idx = cg.PLTC
self.path = 'powerLawTreeGraphs'
elif self.type == 'random_graph':
self.graph = nx.gnm_random_graph(
int(params[0]), int(params[1]),
None if len(params) < 3 else int(params[2]),
False if len(params) < 4 else bool(params[3]))
cg.RGC += 1
self.idx = cg.RGC
self.path = 'randomGraphs'
elif self.type == 'erdos_renyi_random':
self.graph = nx.erdos_renyi_graph(
int(params[0]), float(params[1]),
None if len(params) < 3 else int(params[2]),
False if len(params) < 4 else bool(params[3]))
cg.ERRC += 1
self.idx = cg.ERRC
self.path = 'erdosRenyiRandomGraphs'
else:
print 'GRAPH TYPE:', self.type
raise Exception('Invalid Type of Graph input into argv[2]')
def GenerateRandom(cls,
nodes_nmb,
edges_nmb,
capacity_min=0,
capacity_max=0):
"""
Create graph with specified number of nodes and arcs, where random nodes are connected to random nodes.
Does not contain self loop.
Set to every arc in the graph random value of the capacity from min to max.
"""
network_graph = nx.gnm_random_graph(nodes_nmb,
edges_nmb,
directed=True)
nx.set_edge_attributes(network_graph, {
edge: np.around(
(capacity_max - capacity_min) * nrnd.ranf() + capacity_min, 3)
for edge in nx.edges(network_graph)
},
name='capacity')
self = cls(network_graph)
self.capacity_min = capacity_min
self.capacity_max = capacity_max
return self
def construct_example(n, q, bits = 6):
done = False
m = int(np.ceil(n*np.log(n)/4))
G = nx.gnm_random_graph(n,m)
edges = np.array(G.edges())
# print("edges",edges)
edges = edges.flatten()
queries = np.array([np.random.choice(n, 2, replace=False) for _ in range(q)])
connected = np.zeros(q, dtype=np.float32)
# print("queries",queries)
for qi in range(q):
try:
shortest_paths = np.array([expand_shortest_path(p) for p in
nx.all_shortest_paths(G, queries[qi,0], queries[qi,1])])
connected[qi] = 1.0
except:
connected[qi] = 0.0
# print("shortest_paths",shortest_paths)
# print("length_of_shortest_path",length_of_shortest_path)
# print("connected", connected)
edges = get_bit_sequence(edges, bits)
queries = get_bit_sequence(queries.flatten())
return edges, queries, connected
def CreateRandom(n, m, n_opinion=None, simple_graph=False, seed=None):
"""
Creates an OpinionGraph with an interal graph which is a $G_{n,m}$ random graph, with $n_opinions$ randomly assigned.
In the $G_{n,m}$ model, a graph is chosen uniformly at random from the set
of all graphs with $n$ nodes and $m$ edges.
Parameters
----------
n : int
The number of nodes.
m : int
The number of edges.
n_opinion : int, optional
The number of opinions (default=None, which gives n_opinion=n).
simple_graph : bool, optional
True if the graph is simple (default=None).
seed : int, optional
Seed for random number generator (default=None).
Returns
-------
G : OpinionGraph
"""
internal_graph = nx.gnm_random_graph(n, m)
if not simple_graph:
internal_graph = nx.MultiGraph(internal_graph)
graph = OpinionGraph(internal_graph, n_opinion, simple_graph)
return graph
def losowy_graf(self, n, m, maks_cap, orient):
g = nx.gnm_random_graph(n, m)
graf = [ [] for x in range(0, n)]
# print str(graf)
c = zeros((n, n))
for i in range(0, n):
for j in range(0,n):
if i < j: # gererujemy graf nieskierowany
c[i,j] = random.randint(0, maks_cap)
if orient:
c[j,i] = c[i,j]
else:
c[j,i] = random.randint(0, maks_cap)
if orient:
self.zorientowany = True
#print str(g.edges())
for (u,v) in g.edges():
# print 'Dodajemy wierzcholek'
(graf[u]).append(v)
graf[v].append(u)
#print str(graf)
# for u in range(0,n):
# for v in graf[u]:
# print str(u) + ' -> '+ str(v)
return (graf, c)
def evaluation2(args, original_network):
n = original_network.shape[0]
m = int(np.sum(original_network))
random_ER_G = nx.gnm_random_graph(n, m)
random_bara_G = nx.generators.random_graphs.barabasi_albert_graph(n, 250)
ER_G = nx.to_numpy_matrix(random_ER_G)
bara_G = nx.to_numpy_matrix(random_bara_G)
np.save('{}/ER.npy'.format(args.output_dir), ER_G)
np.save('{}/BA.npy'.format(args.output_dir), bara_G)
# ER_G = ER_G + ER_G .T
# ER_G[ER_G > 1] = 1
# bara_G = bara_G + bara_G .T
# bara_G[bara_G > 1] = 1
ER_G = np.array(ER_G)
bara_G = np.array(bara_G)
print('Computing metrics for ER graph')
c = compute_graph_statistics(ER_G)
with open('./{}/ER_statistics.pickle'.format(args.checkpoint),
'wb') as handle:
pickle.dump(c, handle, protocol=pickle.HIGHEST_PROTOCOL)
print(c)
print('Computing metrics for bara graph')
d = compute_graph_statistics(bara_G)
with open('./{}/bara_statistics.pickle'.format(args.checkpoint),
'wb') as handle:
pickle.dump(d, handle, protocol=pickle.HIGHEST_PROTOCOL)
print(d)
def generate_test ( filenum, isCyclic, nodes, edges, prefix=None ):
while True:
if isCyclic:
G = nx.gnm_random_graph(nodes, edges)
else:
G = nx.random_tree(nodes)
if nx.is_connected(G): # ensure connected graph
break
# nx.draw(G, with_labels=True, font_weight='bold')
# plt.savefig("{}tests/test{}.png".format(prefix if prefix else "",filenum))
# plt.close()
with open("{}tests/test{}.txt".format(prefix if prefix else "",filenum), "w") as infile:
infile.write("{}\n".format(nodes))
A = nx.adjacency_matrix(G).toarray()
for row in A:
for col in row:
infile.write("{} ".format(col))
infile.write("\n")
with open("{}queries/query{}.txt".format(prefix if prefix else "",filenum), "w") as qfile:
source = random.randrange(nodes)
qfile.write("{}\n".format(source))
target = source
while target == source:
target = random.randrange(nodes) # make sure source and target are different nodes
qfile.write("{}\n".format(target))
with open("{}edgelists/edgelist{}.txt".format(prefix if prefix else "",filenum), "wb") as efile:
nx.write_edgelist(G, efile, data=False)
def CriaGrafoAleatorio(n, direct, pond):
# if direct == False:
# Cria um grafo completo nao direcionado
# G = nx.Graph()
# else: # Cria um grafo completo direcionado
# G = nx.DiGraph()
G = nx.gnm_random_graph(n,
random.randint(n - 1,
n * (n - 1) / 2),
seed=time.time(),
directed=direct)
x = random.sample(range(100), n)
y = random.sample(range(100), n)
# Adiciona vertices ao grafo
for i in range(n):
G.add_node(i, pos=(x[i], y[i]))
if pond == True:
# Adiciona arestas com pesos formando um grafo nao direcionado ponderado completo
for e in G.edges():
G.add_edge(e[0],
e[1],
weight=euclid_dist((x[e[0]], x[e[1]]),
(y[e[0]], y[e[1]])))
return G
def generate_random_graph(self):
print 'performer.generate_random_graph:', self.NODE_COUNT, self.EDGE_COUNT
while True:
graph = gnm_random_graph(self.NODE_COUNT, self.EDGE_COUNT)
if self.constraints_satisfied(graph):
self.process_graph(graph)
return graph
def debug_mode(self, param):
G = nx.gnm_random_graph(param['nNodes'], param['nEdges'])
self.modelData = {'nodes': [], 'edges': []}
# Store nodes in self.modelData['nodes']
node_data = [n for n in G.nodes_iter()]
for n in node_data:
tmp = {
'node_id': str(n),
'label': str(n),
'size': random.uniform(param['minNodeSize'],
param['maxNodeSize'])
}
self.modelData['nodes'].append(tmp)
# Store category information
edgeTypes = param['edgeTypes']
num_edgeTypes = len(edgeTypes)
edge_data = list(G.edges_iter(data=False))
for e in edge_data:
tmp = {
'source':
str(e[0]),
'target':
str(e[1]),
'strength':
random.uniform(param['minEdgeStrength'],
param['maxEdgeStrength'])
}
self.modelData['edges'].append(tmp)
return True
def simulate_dag(d, s0):
"""Simulate random DAG with some expected number of edges.
Args:
d (int): num of nodes
s0 (int): expected num of edges
Returns:
B (np.ndarray): [d, d] binary adj matrix of DAG
Function based xunzheng/notears
"""
def _random_permutation(M):
# np.random.permutation permutes first axis only
P = np.random.permutation(np.eye(M.shape[0]))
return P.T @ M @ P
def _random_acyclic_orientation(B_und):
return np.tril(_random_permutation(B_und), k=-1)
G_und = nx.gnm_random_graph(d, s0)
B_und = nx.to_numpy_matrix(G_und)
B = _random_acyclic_orientation(B_und)
B_perm = _random_permutation(B)
assert nx.is_directed_acyclic_graph(
nx.from_numpy_matrix(B_perm, create_using=nx.DiGraph))
return B_perm
def random_graph(N, E):
'''
Creates adjacency list representation of a random graph with N nodes and E edges
'''
G = nx.gnm_random_graph(N, E)
draw_graph(G)
return nx.convert.to_dict_of_dicts(G)
def generate_test(filenum, isTree, nodes, edges, prefix=None):
if isTree:
G = nx.random_tree(nodes)
else:
G = nx.gnm_random_graph(nodes, edges)
# nx.draw(G, with_labels=True, font_weight='bold')
# plt.savefig("{}tests/test{}.png".format(prefix if prefix else "",filenum))
# plt.close()
with open("{}tests/test{}.txt".format(prefix if prefix else "", filenum),
"w") as infile:
infile.write("{}\n".format(nodes))
A = nx.adjacency_matrix(G).toarray()
for row in A:
for col in row:
infile.write("{} ".format(col))
infile.write("\n")
with open(
"{}answers/answer{}.txt".format(prefix if prefix else "", filenum),
"w") as outfile:
try:
cycle = nx.find_cycle(G)
except nx.NetworkXNoCycle:
cycle = None
outfile.write("no" if cycle else "yes")
def create_or_read_simple_graph(name: str) -> nx.Graph:
"""Read a simple unweighted graph from the specified file or create random G_n,m graph if name is gnm-n-m.
The nodes are labeled beginning with 0.
File format:
- ``c <comments> #`` ignored
- ``p <name> <number of nodes> <number of edges>``
- ``e <node_1> <node_2> #`` for each edge, nodes are labeled in 1...number of nodes
"""
if name.startswith('gnm-'):
# create random G_n,m graph
par = name.split(sep='-')
return nx.gnm_random_graph(int(par[1]), int(par[2]), int(par[3]) if len(par) == 4 else None)
else: # read from file
graph: nx.Graph = nx.Graph()
with open(name) as f:
for line in f:
flag = line[0]
if flag == 'p':
split_line = line.split(' ')
n = int(split_line[2])
# m = int(split_line[3])
graph.add_nodes_from(range(n))
elif flag == 'e':
split_line = line.split(' ')
u = int(split_line[1]) - 1
v = int(split_line[2]) - 1
graph.add_edge(u, v)
return graph
def generate_random_graphs(numberOfNodes, outputFile, graphType, degree=None):
sparseResult = open(outputFile, "w")
#first writing the number of nodes
if (graphType == "degreeBased"):
G = nx.random_regular_graph(
degree, numberOfNodes,
numberOfNodes * int(math.sqrt(numberOfNodes)))
if (graphType == "completeChaos"):
G = nx.gnm_random_graph(numberOfNodes,
numberOfNodes * int(math.sqrt(numberOfNodes)))
if (graphType == "dense"):
G = nx.dense_gnm_random_graph(
numberOfNodes, numberOfNodes * int(math.sqrt(numberOfNodes)))
sparseResult.write(
str(numberOfNodes) + " " + str(nx.number_of_edges(G)) + "\n")
semiSparseRep = nx.to_dict_of_lists(G)
#print semiSparseRep
for element in semiSparseRep:
if len(semiSparseRep[element]) == 0:
return 0
for j in semiSparseRep[element]:
sparseResult.write(str(j + 1) + " ")
sparseResult.write("\n")
return 1
def gen_random_networks(n, m, U, num=20):
'''
Input
-----
n : number of nodes in the network
m : number of edges in the network
U : maximum capacity of any arc in the network
num : total number of networks to be generated (defaults to 20)
Output
------
'num' feasible non-trivial networks in forward-star representation
feasible signify networks with s-t paths,
non-trivial signifies networks where there is no direct s-t arc
'''
feasible_networks = []
for i in range(num):
fs_repr = []
isNotFeasible = True
while isNotFeasible:
net = nx.gnm_random_graph(
n, m, directed=True) # Use of networkx library method
if (exists_st_path(net.edges, 0, n - 1)):
isNotFeasible = False
for edge in net.edges:
rc = np.random.randint(
1, U) # generate random capacity of the edge between [1, U]
fs_repr.append([edge[0] + 1, edge[1] + 1,
rc]) # converting 1-based index for the algorithms
#print(fs_repr)
feasible_networks.append(fs_repr)
return feasible_networks
def losowy_graf(self, n, m, maks_cap, orient):
g = nx.gnm_random_graph(n, m)
graf = [[] for x in range(0, n)]
# print str(graf)
c = zeros((n, n))
for i in range(0, n):
for j in range(0, n):
if i < j: # gererujemy graf nieskierowany
c[i, j] = random.randint(0, maks_cap)
if orient:
c[j, i] = c[i, j]
else:
c[j, i] = random.randint(0, maks_cap)
if orient:
self.zorientowany = True
#print str(g.edges())
for (u, v) in g.edges():
# print 'Dodajemy wierzcholek'
(graf[u]).append(v)
graf[v].append(u)
#print str(graf)
# for u in range(0,n):
# for v in graf[u]:
# print str(u) + ' -> '+ str(v)
return (graf, c)
def test_kcores(self):
# Check that results correspond to those of networkx.
G = nx.gnm_random_graph(100, 300)
core_graph = nx.algorithms.k_core(G, 4)
result, _ = k_cores(G, 4)
self.assertEqual(result.edges, core_graph.edges)
def MinDiameterGraphFinder(number_of_vertices, number_of_edges):
if number_of_edges < number_of_vertices - 1:
#print 'Too few edges. The graph will be disconnected. Exiting...'
return -1
if number_of_edges > (number_of_vertices - 1) * number_of_vertices / 2.0:
#print 'Too many edges. Such a graph cannot exist. Exiting...'
return -1
if number_of_edges > (number_of_vertices - 2) * (number_of_vertices -
1) / 2.0:
OutputGraph = nx.gnm_random_graph(number_of_vertices, number_of_edges)
return OutputGraph
G = nx.star_graph(number_of_vertices - 1)
VertexList = G.nodes()
remaining_edges = number_of_edges - number_of_vertices + 1
while remaining_edges > 0:
first_vertex = random.choice(VertexList)
second_vertex = random.choice(VertexList)
if first_vertex == second_vertex: continue
if (first_vertex, second_vertex) in G.edges(): continue
if (second_vertex, first_vertex) in G.edges(): continue
G.add_edge(first_vertex, second_vertex)
remaining_edges -= 1
OutputGraph = G.copy()
return OutputGraph
def generate_graph(n, m, seed):
G = nx.gnm_random_graph(n, m, seed)
data = json_graph.node_link_data(G)
with open('./Working/graph.json', 'w') as f:
json.dump(data, f)
with open('./Working/graph_flag', mode = 'w', encoding = 'utf-8') as fh:
fh.write("i look at you")
def Densityrandom_date(density):
density=density/100
nodes=random.randint(3,200)
edges= (density*nodes*(nodes-1))//2
print(density,"nodes",nodes,"edges", edges)
domain = {}
constraint = {}
G = nx.gnm_random_graph(nodes, edges, False)
# print(G.nodes)
# print(G.edges)
for nd in G.nodes:
# domain[nd]=random.sample(range(1, 100), 10)
domain[nd] = random.sample(range(1, 1000), 50)
numberOfConst = 6
for e in G.edges:
constraint[e] = random.randint(0, numberOfConst)
# print(domain)
# print(constraint)
return (G, domain, constraint)
def main():
args = parse_arguments()
# Generate graph for given parameters
if args.cycle:
graph = networkx.cycle_graph(args.vertices)
graph = networkx.path_graph(args.vertices)
elif args.star:
graph = networkx.star_graph(args.vertices)
elif args.wheel:
graph = networkx.wheel_graph(args.vertices)
elif args.ladder:
graph = networkx.ladder_graph(args.vertices)
elif args.fill_rate:
if args.fill_rate > 50.0:
graph = networkx.gnp_random_graph(args.vertices, args.fill_rate / 100.0, None, args.directed)
else:
graph = networkx.fast_gnp_random_graph(args.vertices, args.fill_rate / 100.0, None, args.directed)
else:
graph = networkx.gnm_random_graph(args.vertices, args.edges, None, args.directed)
# Print generated graph
print("{} {}".format(graph.number_of_nodes(), graph.number_of_edges()))
for edge in graph.edges():
print("{} {}".format(edge[0] + 1, edge[1] + 1))
# Export generated graph to .png
if args.image_path:
export_to_png(graph, args.image_path)
def test():
import networkx as nx
network = nx.gnm_random_graph(50,
50,
directed=True)
sim = Simulation(10, 50, network)
# generate users
print('Generation')
print('=' * 20)
for n in network.nodes():
u1 = User(n, sim.topics)
sim.add_user(u1)
print(sim.get_user(n))
# add follow
print('Follow')
print('=' * 20)
for e in network.edges():
sim.new_follow(e[0],e[1])
print(sim.get_user(e[0]).get_attachment(e[1]))
print('Simulation')
print('=' * 20)
flag = 0
while(flag>-1):
print('Time: %d Edges: %d' % (sim.now, len(network.edges())))
print('Tweets: %d' % len(sim.tweet))
print('Retweets: %d' % len(sim.retweet))
sim.step_tweet()
sim.step_retweet()
sim.now += 1
flag += 1
def ErdosRenyi(n, m, d, display=None, seed=None):
"""
n: the number of nodes
m: the number of edges
"""
# counter = 0
# naive way to generate connected graph
while True:
if m <= 30:
G = nx.gnm_random_graph(n, m, seed=None, directed=False)
else:
G = nx.dense_gnm_random_graph(n, m, seed=None)
maxDegree = max([item for item in G.degree().values()])
if nx.is_connected(G) and maxDegree <= d:
break
# if maxDegree > d:
# counter += 1
# avgDegree = 2 * nx.number_of_edges(G) / nx.number_of_nodes(G)
# print avgDegree
if display:
nx.draw(G)
img_name = 'erdos_renyi_%i_%i.png' % (n, m)
plt.savefig(img_name)
sh.open(img_name)
adjacencyMatrix = generateAdjacencyMatrix(G)
return adjacencyMatrix
def random_graph(self):
#Compute graph and caculate average path length, clustering coeff
logging.info('Inside random graph')
RG = nx.gnm_random_graph(self.node_count, self.edge_count)
RG = nx.connected_component_subgraphs(RG.to_undirected())[0]
self.average_shortest_path_length('Random graph', RG)
self.global_clustering_coefficient('Random graph', RG)
def connect_fixed_edges(self,N,p):
"""
A fixed fraction of the total possible N*(N-1)/2 connections are made. (Not a binomial
graph! The number of edges is always p*N(N-1)/2, not just in the N->infinity limit.)
"""
self.connect_empty(N)
dN = int(p*N*(N-1)/2)
self.add_edges_from(nx.gnm_random_graph(N,dN).edges())
def random_graph_heterogeneous_synapses(cell_positions):
h = spatial_graph_2010(cell_positions)
weights = [e[2]['weight'] for e in h.edges(data=True)]
random.shuffle(weights)
g = nx.gnm_random_graph(h.order(), h.size())
for k, e in enumerate(g.edges()):
g[e[0]][e[1]]['weight'] = weights[k]
return g
def obtain_graph(args):
"""Build a Graph according to command line arguments
Arguments:
- `args`: command line options
"""
if hasattr(args,'gnd') and args.gnd:
n,d = args.gnd
if (n*d)%2 == 1:
raise ValueError("n * d must be even")
G=networkx.random_regular_graph(d,n)
return G
elif hasattr(args,'gnp') and args.gnp:
n,p = args.gnp
G=networkx.gnp_random_graph(n,p)
elif hasattr(args,'gnm') and args.gnm:
n,m = args.gnm
G=networkx.gnm_random_graph(n,m)
elif hasattr(args,'grid') and args.grid:
G=networkx.grid_graph(args.grid)
elif hasattr(args,'torus') and args.torus:
G=networkx.grid_graph(args.torus,periodic=True)
elif hasattr(args,'complete') and args.complete>0:
G=networkx.complete_graph(args.complete)
elif args.graphformat:
G=readGraph(args.input,args.graphformat)
else:
raise RuntimeError("Invalid graph specification on command line")
# Graph modifications
if hasattr(args,'plantclique') and args.plantclique>1:
clique=random.sample(G.nodes(),args.plantclique)
for v,w in combinations(clique,2):
G.add_edge(v,w)
# Output the graph is requested
if hasattr(args,'savegraph') and args.savegraph:
writeGraph(G,
args.savegraph,
args.graphformat,
graph_type='simple')
return G
def demo():
'''
Run a demo showing how a random, ugly network graph evolves into
a much nicer-looking graph.
'''
size = 8
G = nx.gnm_random_graph(size,11)
seed = nx.spring_layout(G)
evol = run_simulation(G,seed)
def graphFactoryPlanarNormalDistribution(nb_graph,size_graph,graphtype,location,spread,section='all'):
cpt=0
while cpt <= nb_graph:
rdm_density = np.random.normal(location,spread)
G = nx.gnm_random_graph(size_graph,edgeForDensity(size_graph,rdm_density))
if graphToCSV(G,graphtype,section,pl.is_planar(G)):
cpt+=1
if cpt%10 == 0:
print(str(cpt)+'/'+str(nb_graph)+' '+str(100*cpt/nb_graph)+'%')
def graphFactoryPlanar(nb_graph, size_graph, graphtype, section='all'):
cpt = 0
while cpt <= nb_graph:
m = np.random.random_sample(1)
G = nx.gnm_random_graph(size_graph,edgeForDensity(size_graph,m))
if graphToCSV(G,graphtype,section,pl.is_planar(G)):
cpt+=1
if cpt%10 == 0:
print(str(cpt)+'/'+str(nb_graph)+' '+str(100*cpt/nb_graph)+'%')
def build_random_network(n=100, m=300, directed=False):
"""
Builds a random network with n nodes (default n=100) and m edges
(default m=300).
"""
network = nx.gnm_random_graph(n, m, directed=directed)
return network
def RandomGNM(n, m, dense=False, seed=None):
"""
Returns a graph randomly picked out of all graphs on n vertices
with m edges.
INPUT:
- ``n`` - number of vertices.
- ``m`` - number of edges.
- ``dense`` - whether to use NetworkX's
dense_gnm_random_graph or gnm_random_graph
EXAMPLES: We show the edge list of a random graph on 5 nodes with
10 edges.
::
sage: graphs.RandomGNM(5, 10).edges(labels=False)
[(0, 1), (0, 2), (0, 3), (0, 4), (1, 2), (1, 3), (1, 4), (2, 3), (2, 4), (3, 4)]
We plot a random graph on 12 nodes with m = 12.
::
sage: gnm = graphs.RandomGNM(12, 12)
sage: gnm.show() # long time
We view many random graphs using a graphics array::
sage: g = []
sage: j = []
sage: for i in range(9):
... k = graphs.RandomGNM(i+3, i^2-i)
... g.append(k)
...
sage: for i in range(3):
... n = []
... for m in range(3):
... n.append(g[3*i + m].plot(vertex_size=50, vertex_labels=False))
... j.append(n)
...
sage: G = sage.plot.graphics.GraphicsArray(j)
sage: G.show() # long time
"""
if seed is None:
seed = current_randstate().long_seed()
import networkx
if dense:
return graph.Graph(networkx.dense_gnm_random_graph(n, m, seed=seed))
else:
return graph.Graph(networkx.gnm_random_graph(n, m, seed=seed))
def setUp(self):
self.graph = nx.gnm_random_graph(N,E)
for e in self.graph.edges():
self.graph[e[0]][e[1]]['weight'] = random.random()
self.atheta = { n : np.random.random((Z))/Z for n in self.graph.nodes() }
pcgpath = cg_path + 'model.social.tapmodel.TAPModel.__init__.png'
with Profile(pcgpath):
self.T = TAPModel(self.graph, self.atheta)
def demo_evolution():
'''
Run a demo showing how a random, ugly network graph evolves into
a much nicer-looking graph, using an asexually-reproducing
evolutionary algorithm.
'''
size = 8
G = nx.gnm_random_graph(size,11)
seed = nx.spring_layout(G)
evol = evolve_asexually(G,seed)
def make_graph(num_nodes, num_edges, seed=None, directed=True):
# create the graph
# (if we use the same seed every time, we'll get the same graph)
graph = nx.gnm_random_graph(num_nodes, num_edges, seed=seed,
directed=directed)
# assign edge weights
# (again we use a seed so we get the same weights every time)
random.seed(seed)
for _, _, edge_attrs in graph.edges_iter(data=True):
edge_attrs['weight'] = random.randint(0, 99)
return graph
def create_3(nodes, edges):
conn = 1
loop_count = 0
while conn < 3 and loop_count < 1000:
G = nx.gnm_random_graph(nodes, edges)
conn = nx.node_connectivity(G)
loop_count = loop_count + 1
return G
def random_connected_graph(numNodes, numEdges):
'''Generates a weakly connected random graph with numNodes nodes
and numEdges edges
'''
tries = 0
while tries < 100:
tries += 1
random_graph = nx.gnm_random_graph(numNodes,numEdges,directed = True)
if nx.is_weakly_connected(random_graph):
return random_graph
return None
def randomize_graph(n,e):
"""
Creates 1000 random graphs with networkx.gnm_random_graph(nodecount, edgecount),
and computes average_clustering_coefficient and average_shortest_path_length,
to compare with drama-graph.
Returns a tuple:
randavgpathl, randcluster = (float or "NaN", float or "NaN")
"""
randcluster = 0
randavgpathl = 0
c = 0
for i in range(0, 1000):
R = nx.gnm_random_graph(n, e)
try:
randcluster += nx.average_clustering(R)
c += 1
except ZeroDivisionError:
pass
j = 0
while True:
j += 1
try:
R = nx.gnm_random_graph(n, e)
randavgpathl += nx.average_shortest_path_length(R)
except:
pass
else:
break
if j > 50:
randavgpathl = "NaN"
break
try:
randcluster = randcluster / c
except:
randcluster = "NaN"
try:
randavgpathl = randavgpathl / 1000
except:
randavgpathl = "NaN"
return randavgpathl, randcluster
def make_gnms():
for n in [10, 1000, 2000, 3000, 4000, 5000]:
for d in [3,4]:
m = (d*n)/2
G = nx.gnm_random_graph(n,m)
R = random.sample(G.nodes(), n/10)
name = 'gnm-{0}-{1}-0'.format(n,m)
write_graph(G, name, 0, 1, R)
R = random.sample(G.nodes(), n/2)
name = 'gnm-{0}-{1}-1'.format(n,m)
write_graph(G, name, 0, 1, R)
def _generate_graph(self):
samples = [
nx.gnm_random_graph(self._cstarrt['Nodes'], self._cst['Edges'], directed=True),
]
while self._maxtry:
self._maxtry -= 1
graph = random.sample(samples, 1)[0]
if self._is_valid_graph(graph):
return graph
print "can't generate DAG worker graph"
raise
def random_regulatory(num_nodes, num_activating, num_inhibiting,
function="function", seed=None):
net = nx.gnm_random_graph(num_nodes, num_activating + num_inhibiting,
seed=seed, directed=True)
edges = net.edges()
activating = random.sample(edges, num_activating)
for (u, v) in activating:
net[u][v][function] = 1
for (u, v) in list(set(edges) - set(activating)):
net[u][v][function] = -1
add_self_inhibition(net, function)
return net
def obtain_graph(args,suffix=""):
"""Build a Graph according to command line arguments
Arguments:
- `args`: command line options
"""
if getattr(args,'gnd'+suffix) is not None:
n,d = getattr(args,'gnd'+suffix)
if (n*d)%2 == 1:
raise ValueError("n * d must be even")
G=networkx.random_regular_graph(d,n)
elif getattr(args,'gnp'+suffix) is not None:
n,p = getattr(args,'gnp'+suffix)
G=networkx.gnp_random_graph(n,p)
elif getattr(args,'gnm'+suffix) is not None:
n,m = getattr(args,'gnm'+suffix)
G=networkx.gnm_random_graph(n,m)
elif getattr(args,'grid'+suffix) is not None:
G=networkx.grid_graph(getattr(args,'grid'+suffix))
elif getattr(args,'torus'+suffix) is not None:
G=networkx.grid_graph(getattr(args,'torus'+suffix),periodic=True)
elif getattr(args,'complete'+suffix) is not None:
G=networkx.complete_graph(getattr(args,'complete'+suffix))
elif getattr(args,'empty'+suffix) is not None:
G=networkx.empty_graph(getattr(args,'empty'+suffix))
elif getattr(args,'graphformat'+suffix) is not None:
try:
print("INFO: reading simple graph {} from '{}'".format(suffix,getattr(args,"input"+suffix).name),
file=sys.stderr)
G=readGraph(getattr(args,'input'+suffix),
"simple",
getattr(args,'graphformat'+suffix))
except ValueError,e:
print("ERROR ON '{}'. {}".format(
getattr(args,'input'+suffix).name,e),
file=sys.stderr)
exit(-1)
def __init__(self, n, m, seed=None):
""" Produces a graph picked randomly
out of the set of all graphs
with n nodes and m edges.
@param n: Number of nodes
@param m: Number of edges
@param seed: Seed for random number generator (default=None).
@return: The Topology
"""
Topo.__init__(self, name='Random')
self.n = n
self.m = m
self.G = nx.gnm_random_graph(n, m)