def run_example():
network_graph = load_network_graph()
num_nodes = len(network_graph)
print "Total number of nodes in Network Graph : ", num_nodes
total_edges = 0
for node in network_graph:
edge = network_graph[node]
total_edges += len(edge)
total_edges /= 2
print "Total number of edges in Network Graph : ", total_edges
prob_edge = float((total_edges * 2)) / (num_nodes * (num_nodes - 1))
er_graph = load_er_graph(num_nodes, prob_edge)
print "Total number of nodes in ER Graph : ", num_nodes
total_edges = 0
for node in er_graph:
edge = er_graph[node]
total_edges += len(edge)
total_edges /= 2
print "Total number of edges in ER Graph : ", total_edges, " with probability : ", prob_edge
approx_m = total_edges / num_nodes
exact_m = int(
math.ceil(0.5 *
(-math.sqrt(((2 * num_nodes - 1)**2) - 8 * total_edges) +
(2 * num_nodes - 1))))
upa_graph = load_upa_graph(num_nodes, approx_m)
print "Total number of nodes in UPA Graph : ", num_nodes
total_edges = 0
for node in upa_graph:
edge = upa_graph[node]
total_edges += len(edge)
total_edges /= 2
print "Total number of edges in UPA Graph : ", total_edges, " with m : ", approx_m
attack_network = fast_targeted_order(network_graph)
resilience_network = cc_gr.compute_resilience(
network_graph, attack_network[:len(attack_network) / 5])
attack_er = fast_targeted_order(er_graph)
resilience_er = cc_gr.compute_resilience(er_graph,
attack_er[:len(attack_er) / 5])
attack_upa = fast_targeted_order(upa_graph)
resilience_upa = cc_gr.compute_resilience(upa_graph,
attack_upa[:len(attack_upa) / 5])
print resilience_network[-1], len(network_graph)
print resilience_er[-1], len(er_graph)
print resilience_upa[-1], len(upa_graph)
def run_example():
network_graph = load_network_graph()
num_nodes = len(network_graph)
print "Total number of nodes in Network Graph : ", num_nodes
total_edges = 0
for node in network_graph:
edge = network_graph[node]
total_edges += len(edge)
total_edges /= 2
print "Total number of edges in Network Graph : ", total_edges
prob_edge = float((total_edges * 2)) / (num_nodes * (num_nodes - 1))
er_graph = load_er_graph(num_nodes, prob_edge)
print "Total number of nodes in ER Graph : ", num_nodes
total_edges = 0
for node in er_graph:
edge = er_graph[node]
total_edges += len(edge)
total_edges /= 2
print "Total number of edges in ER Graph : ", total_edges, " with probability : ", prob_edge
approx_m = total_edges / num_nodes
exact_m = int(math.ceil(0.5 * (-math.sqrt(((2 * num_nodes - 1) ** 2) - 8 * total_edges) + (2 * num_nodes - 1))))
upa_graph = load_upa_graph(num_nodes, approx_m)
print "Total number of nodes in UPA Graph : ", num_nodes
total_edges = 0
for node in upa_graph:
edge = upa_graph[node]
total_edges += len(edge)
total_edges /= 2
print "Total number of edges in UPA Graph : ", total_edges, " with m : ", approx_m
attack_network = random_order(network_graph)
resilience_network = cc_gr.compute_resilience(network_graph, attack_network[:len(attack_network)/5])
attack_er = random_order(er_graph)
resilience_er = cc_gr.compute_resilience(er_graph, attack_er[:len(attack_er)/5])
attack_upa = random_order(upa_graph)
resilience_upa = cc_gr.compute_resilience(upa_graph, attack_upa[:len(attack_upa)/5])
print resilience_network[-1], len(network_graph)
print resilience_er[-1], len(er_graph)
print resilience_upa[-1], len(upa_graph)
def question1():
"""
In question 1, we will examine the resilience of the computer network
under an attack in which servers are chosen at random. We will then compare
the resilience of the network to the resilience of ER and UPA graphs of
similar size.
"""
# Determine the probability such that the ER graph has approx. the same
# number of edges as the computer network. Likewise, compute an integer such
# that the number of edges in the UPA graph is close to the number of edges
# in the computer network. Each graph should have the same number of nodes, 1239.
ER_prob = .004
UPA_num_existing_nodes = 3
ER_graph = make_ER_graph(1239, ER_prob)
UPA_graph = make_UPA_graph(1239, UPA_num_existing_nodes)
network_graph = make_computer_network_graph()
# Compute random attack order for each graph.
ER_random = random_order(ER_graph)
UPA_random = random_order(UPA_graph)
network_random = random_order(network_graph)
# Compute the resilience of each graph.
ER_resilience = project2.compute_resilience(ER_graph, ER_random)
UPA_resilience = project2.compute_resilience(UPA_graph, UPA_random)
network_resilience = project2.compute_resilience(network_graph, network_random)
# Plot the results
xvals = range(len(network_resilience))
ER_yvals = ER_resilience
UPA_yvals = UPA_resilience
network_yvals = network_resilience
plt.title("Question 1: Resilience of Computer Network, ER Graph, and UPA Graph")
plt.xlabel("Number of Nodes Removed")
plt.ylabel("Size of Largest Connected Component")
plt.plot(xvals, ER_yvals, '-b', label='ER Graph (p=.004)')
plt.plot(xvals, UPA_yvals, '-r', label='UPA Graph (m=3)')
plt.plot(xvals, network_yvals, '-g', label='Computer Network')
plt.grid(which='major', axis='both')
plt.legend(loc='upper right')
plt.show()
def question4():
"""
This function:
1) Uses fast_targeted_order to compute a targeted attack order for
each of the three graphs (computer network, ER, UPA) from Question 1.
2) Uses compute_resilience to compute the resilience of each graph.
3) Plots the computed resiliences as three curves in a single standard plot.
"""
# Three graphs. The probability argument (.004) for make_ER_graph and
# num_existing_nodes argument (3) for make_UPA_graph are same as in question1.
network_graph = make_computer_network_graph()
ER_graph = make_ER_graph(1239, .004)
UPA_graph = make_UPA_graph(1239, 3)
# Computes a targeted attack order for each of the three graphs.
network_order = fast_targeted_order(network_graph)
ER_order = fast_targeted_order(ER_graph)
UPA_order = fast_targeted_order(UPA_graph)
# Computes graph resilience for each of the three graphs.
network_resilience = project2.compute_resilience(network_graph, network_order)
ER_resilience = project2.compute_resilience(ER_graph, ER_order)
UPA_resilience = project2.compute_resilience(UPA_graph, UPA_order)
# Plot the computed resiliences.
xvals = range(len(network_resilience))
ER_yvals = ER_resilience
UPA_yvals = UPA_resilience
network_yvals = network_resilience
plt.title("Question 4: Graph and Network Resilience Under Connectivity-Based Attack")
plt.xlabel("Number of Nodes Removed")
plt.ylabel("Size of Largest Connected Component")
plt.plot(xvals, ER_yvals, '-b', label='ER Graph (p=.004)')
plt.plot(xvals, UPA_yvals, '-r', label='UPA Graph (m=3)')
plt.plot(xvals, network_yvals, '-g', label='Computer Network')
plt.grid(which='major', axis='both')
plt.legend(loc='upper right')
plt.show()
def test_compute_resilience():
resilience = cc.compute_resilience(graphs.GRAPH2, [1, 3, 5, 7, 2, 4, 6, 8])
print resilience
def test_compute_resilience():
resilience = cc.compute_resilience(graphs.GRAPH2, [1, 3, 5, 7, 2, 4, 6, 8])
print resilience
