def plot_graph(graph, title='', xlabel='', ylabel=''):
"""
Function uses matplotlib to create a log/log plot of the
in-degree distribution of a graph.
Args:
graph: a dictionary representation of a directed graph
where the keys are node names and the values are sets
of edges.
Returns:
None and plots a log/log plot of the graph.
"""
# Calculate the in-degree distribution of the graph
in_degree_dist = g.in_degree_distribution(graph)
# Normalize the in-degree distribution to sum up to 1
normalized_distribution = g.normalize_distribution(in_degree_dist)
# Plot the normalized distribution on a graph
#plt.plot(normalized_distribution.keys(), normalized_distribution.values(), 'ro')
plt.loglog(normalized_distribution.keys(), normalized_distribution.values(), 'bo', basex=10, basey=10)
# Add attributes to the plot
plt.title(title)
plt.xlabel(xlabel)
plt.ylabel(ylabel)
plt.show()
def plot_graph(graph, title='', xlabel='', ylabel=''):
"""
Function uses matplotlib to create a log/log plot of the
in-degree distribution of a graph.
Args:
graph: a dictionary representation of a directed graph
where the keys are node names and the values are sets
of edges.
Returns:
None and plots a log/log plot of the graph.
"""
# Calculate the in-degree distribution of the graph
in_degree_dist = g.in_degree_distribution(graph)
# Normalize the in-degree distribution to sum up to 1
normalized_distribution = g.normalize_distribution(in_degree_dist)
# Plot the normalized distribution on a graph
#plt.plot(normalized_distribution.keys(), normalized_distribution.values(), 'ro')
plt.loglog(normalized_distribution.keys(),
normalized_distribution.values(),
'bo',
basex=10,
basey=10)
# Add attributes to the plot
plt.title(title)
plt.xlabel(xlabel)
plt.ylabel(ylabel)
plt.show()
def plot_citation_graph():
"""
Function for plotting the in-degree distribution of
a specific citation graph.
"""
# Load the external graph of citation information
citation_graph = g.load_graph(CITATION_URL)
# Calculate the in-degree distribution of the citation graph
in_degree_dist = g.in_degree_distribution(citation_graph)
# Normalize the in-degree distribution to sum up to 1
normalized_distribution = g.normalize_distribution(in_degree_dist)
# Plot the normalized distribution on a log/log graph
plt.loglog(normalized_distribution.keys(),
normalized_distribution.values(),
'ro',
basex=10,
basey=10)
plt.title(
'Log/log plot of the normalized distribution of a citation graph')
plt.ylabel('Number of citations - base 10')
plt.xlabel('Papers - base 10')
plt.show()
def plot_citation_graph():
"""
Function for plotting the in-degree distribution of
a specific citation graph.
"""
# Load the external graph of citation information
citation_graph = g.load_graph(CITATION_URL)
# Calculate the in-degree distribution of the citation graph
in_degree_dist = g.in_degree_distribution(citation_graph)
# Normalize the in-degree distribution to sum up to 1
normalized_distribution = g.normalize_distribution(in_degree_dist)
# Plot the normalized distribution on a log/log graph
plt.loglog(normalized_distribution.keys(), normalized_distribution.values(), 'ro', basex=10, basey=10)
plt.title('Log/log plot of the normalized distribution of a citation graph')
plt.ylabel('Number of citations - base 10')
plt.xlabel('Papers - base 10')
plt.show()
import graph
def er(n, p):
outgraph = dict(enumerate(range(n)))
for key in outgraph.keys():
outgraph[key] = set([])
nodes = set([item for item in outgraph.keys() if item != key])
for node in nodes:
rand = random.random()
if rand < p:
outgraph[key].add(node)
return outgraph
percentage = 0.01
num_nodes = 27770
in_degrees = graph.in_degree_distribution(er(num_nodes, percentage))
def divide(x):
return x / float(num_nodes)
normalized = map(divide, in_degrees.values())
pyplot.loglog(in_degrees.keys(), normalized, 'bo')
pyplot.xlabel('degree')
pyplot.ylabel('normalized frequency')
pyplot.title('ER graph - normalized degree distribution')
pyplot.show()
def dpa(n, m):
outgraph = graph.make_complete_graph(m)
dpat = DPATrial.DPATrial(m)
i = m
while i < n:
outgraph[i] = dpat.run_trial(m)
i += 1
return outgraph
num_nodes = 100
avg_out_degree = 13
dpaGraph = dpa(num_nodes, avg_out_degree)
in_degrees = graph.in_degree_distribution(dpaGraph)
# def divide(x):
# return x / float(num_nodes)
#
# normalized = map(divide, in_degrees.values())
#
#
#
# pyplot.loglog(in_degrees.keys(), normalized, 'bo')
# pyplot.xlabel('degree')
# pyplot.ylabel('normalized frequency')
# pyplot.title('DPA graph - normalized degree distribution')
# pyplot.show()
# -*- coding: utf-8 -*-
"""
Created on Tue Aug 26 21:00:06 2014
@author: Administrator
"""
import graph
import random
import matplotlib.pyplot as plt
def upa(n,m):
result_graph = graph.make_complete_graph(m)
for i in range(m,n):
edeg = []
for count in range(m):
edeg.append(random.choice(result_graph.keys()))
result_graph[i] = set(edeg)
return result_graph
if __name__ == "__main__":
g = upa(27770,13)
edeg_num = 0
for key in g.keys():
edeg_num += len(g[key])
distribution = graph.in_degree_distribution(g)
for dummy_key in distribution.keys():
distribution[dummy_key] = (float)(distribution[dummy_key]) / 27770
plt.loglog(distribution.keys(),distribution.values())
node = int(neighbors[0])
answer_graph[node] = set([])
for neighbor in neighbors[1 : -1]:
answer_graph[node].add(int(neighbor))
return answer_graph
citation_graph = load_graph(CITATION_URL)
total = float(27770)
def divide(x):
""" division method """
return x / total;
in_degrees = graph.in_degree_distribution(citation_graph)
print(in_degrees)
print graph.avg_in_degree(in_degrees)
# normalized = map(divide, in_degrees.values())
# print(len(normalized), "length")
#
# def add(x,y): return x + y
#
# print(reduce(add, normalized))
#
# print normalized
#
# pyplot.loglog(in_degrees.keys(), normalized, 'bo')
# pyplot.xlabel('degree')
def dpa(n, m):
outgraph = graph.make_complete_graph(m)
dpat = DPATrial.DPATrial(m)
i = m
while i < n:
outgraph[i] = dpat.run_trial(m)
i += 1
return outgraph
num_nodes = 100
avg_out_degree = 13
dpaGraph= dpa(num_nodes, avg_out_degree)
in_degrees = graph.in_degree_distribution(dpaGraph)
# def divide(x):
# return x / float(num_nodes)
#
# normalized = map(divide, in_degrees.values())
#
#
#
# pyplot.loglog(in_degrees.keys(), normalized, 'bo')
# pyplot.xlabel('degree')
# pyplot.ylabel('normalized frequency')
# pyplot.title('DPA graph - normalized degree distribution')
# pyplot.show()