initial public release

Author: jayavanth

__init__.py

@@ -0,0 +1,39 @@
+import random 
+import networkx as nx
+
+
+# generate a random graph with degree between a and b
+def generate_graph(degree_start=10, degree_end=15, size=5000):
+
+    z=[random.randint(degree_start, degree_end) for i in range(size)]
+
+    # if sum of degree sequence is odd, make it even
+    if sum(z) % 2 != 0:
+       z[random.randint(0, size)] += 1 
+
+    G=nx.configuration_model(z)
+
+    # remove self-loops and parallel edges
+    G=nx.Graph(G)
+    G.remove_edges_from(G.selfloop_edges())
+
+    return G
+
+
+
+def generate_graph_mean(degree_start=10, degree_end=15, size=5000):
+
+    z=[numpy.random.geometric(0.07) for i in range(size)]
+
+    print("Mean degree", numpy.mean(z))
+    # if sum of degree sequence is odd, make it even
+    if sum(z) % 2 != 0:
+       z[random.randint(0, size)] += 1
+
+    G=nx.configuration_model(z)
+
+    # remove self-loops and parallel edges
+    G=nx.Graph(G)
+    G.remove_edges_from(G.selfloop_edges())
+
+    return G

gen.py

@@ -0,0 +1,51 @@
+import csv
+import matplotlib
+import numpy as np
+
+from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+
+csv_read = csv.reader(open('log.txt'), delimiter=',')
+
+avgd = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]} 
+two3 = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+for row in csv_read:
+    #import ipdb ; ipdb.set_trace()
+    for i, col in enumerate(row):
+        
+        # 0 - <>
+        # 1 - M
+        # 2 - N
+        # 3 - size
+        # 4 - degree min
+        # 5 - degree max
+        # 6 - shortestpath
+        # 7 - 2/3
+        sz = int(row[3])
+        avgd_rw = np.mean([int(row[4]), int(row[5])])
+        two3_rw = float(row[7])
+
+        avgd[sz].append(avgd_rw)
+        two3[sz].append(two3_rw)
+
+
+
+fig = plt.figure()
+ax = fig.add_subplot(111)
+
+ax.scatter(avgd[1000], two3[1000], c='r', marker='o', label='1000')
+ax.scatter(avgd[2000], two3[2000], c='g', marker='1', label='2000')
+ax.scatter(avgd[3000], two3[3000], c='b', marker='2', label='3000')
+ax.scatter(avgd[4000], two3[4000], c='y', marker='x', label='4000')
+ax.scatter(avgd[5000], two3[5000], c='c', marker='+', label='5000')
+
+#ax.set_xscale('log')
+#ax.set_xticklabels([1,2,3,10,100])
+#ax.set_xticks([1,2,3,10,100])
+
+ax.set_xlabel('Average degree')
+ax.set_ylabel('Hops to reach 2/3rds of nodes')
+
+ax.legend()
+
+plt.show()

plot_avg_23.py

@@ -0,0 +1,55 @@
+import csv
+import matplotlib
+import numpy as np
+
+from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+
+csv_read = csv.reader(open('log.txt'), delimiter=',')
+
+m = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+avgd = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]} 
+sp = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+for row in csv_read:
+    #import ipdb ; ipdb.set_trace()
+    for i, col in enumerate(row):
+
+        # 1 - M
+        # 2 - N
+        # 3 - size
+        # 4 - degree min
+        # 5 - degree max
+        # 6 - shortestpath
+        # 7 - 2/3
+        sz = int(row[3])
+        m_rw = int(row[1])
+        avgd_rw = np.mean([int(row[4]), int(row[5])])
+        sp_rw = float(row[6])
+
+        m[sz].append(m_rw)
+        avgd[sz].append(avgd_rw)
+        sp[sz].append(sp_rw)
+
+
+
+
+
+
+
+fig = plt.figure()
+ax = fig.add_subplot(111, projection='3d')
+
+ax.scatter(m[1000], avgd[1000], sp[1000], c='r', marker='x', label='1000')
+ax.scatter(m[2000], avgd[2000], sp[2000], c='g', marker='+', label='2000')
+ax.scatter(m[3000], avgd[3000], sp[3000], c='b', marker='1', label='3000')
+ax.scatter(m[4000], avgd[4000], sp[4000], c='purple', marker='2', label='4000')
+ax.scatter(m[5000], avgd[5000], sp[5000], c='c', marker='3', label='5000')
+
+
+ax.set_xlabel('M')
+ax.set_ylabel('Average degree')
+ax.set_zlabel('Shortest path time')
+
+ax.legend()
+
+plt.show()

plot_m_avgd_sp.py

@@ -0,0 +1,51 @@
+import csv
+import matplotlib
+import numpy as np
+
+from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+
+csv_read = csv.reader(open('log.txt'), delimiter=',')
+
+n = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+two3 = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+for row in csv_read:
+    #import ipdb ; ipdb.set_trace()
+    for i, col in enumerate(row):
+        
+        # 0 - <>
+        # 1 - M
+        # 2 - N
+        # 3 - size
+        # 4 - degree min
+        # 5 - degree max
+        # 6 - shortestpath
+        # 7 - 2/3
+        sz = int(row[3])
+        n_rw = int(row[2])
+        two3_rw = float(row[7])
+
+        n[sz].append(n_rw)
+        two3[sz].append(two3_rw)
+
+
+
+fig = plt.figure()
+ax = fig.add_subplot(111)
+
+ax.scatter(n[1000], two3[1000], c='r', marker='+', label='1000')
+ax.scatter(n[2000], two3[2000], c='g', marker='x', label='2000')
+ax.scatter(n[3000], two3[3000], c='b', marker='1', label='3000')
+ax.scatter(n[4000], two3[4000], c='purple', marker='2', label='4000')
+ax.scatter(n[5000], two3[5000], c='c', marker='3', label='5000')
+
+ax.set_xscale('log')
+ax.set_xticklabels([1,2,3,10,100])
+ax.set_xticks([1,2,3,10,100])
+
+ax.set_xlabel('N')
+ax.set_ylabel('Time to reach 2/3rds of nodes')
+
+ax.legend()
+
+plt.show()

plot_n_23.py

@@ -0,0 +1,51 @@
+import csv
+import matplotlib
+import numpy as np
+
+from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+
+csv_read = csv.reader(open('log.txt'), delimiter=',')
+
+n = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+avgd = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]} 
+for row in csv_read:
+    #import ipdb ; ipdb.set_trace()
+    for i, col in enumerate(row):
+        
+        # 0 - <>
+        # 1 - M
+        # 2 - N
+        # 3 - size
+        # 4 - degree min
+        # 5 - degree max
+        # 6 - shortestpath
+        # 7 - 2/3
+        sz = int(row[3])
+        n_rw = int(row[2])
+        avgd_rw = np.mean([int(row[4]), int(row[5])])
+
+        n[sz].append(n_rw)
+        avgd[sz].append(avgd_rw)
+
+
+
+fig = plt.figure()
+ax = fig.add_subplot(111)
+
+ax.scatter(n[1000], avgd[1000], c='r', marker='o', label='1000')
+ax.scatter(n[2000], avgd[2000], c='g', marker='o', label='2000')
+ax.scatter(n[3000], avgd[3000], c='b', marker='o', label='3000')
+ax.scatter(n[4000], avgd[4000], c='y', marker='o', label='4000')
+ax.scatter(n[5000], avgd[5000], c='c', marker='o', label='5000')
+
+ax.set_xscale('log')
+ax.set_xticklabels([1,2,3,10,100])
+ax.set_xticks([1,2,3,10,100])
+
+ax.set_xlabel('N')
+ax.set_ylabel('Average degree')
+
+ax.legend()
+
+plt.show()

plot_n_avg.py

@@ -0,0 +1,59 @@
+import csv
+import matplotlib
+import numpy as np
+
+from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+
+csv_read = csv.reader(open('log.txt'), delimiter=',')
+
+n = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+avgd = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]} 
+two3 = {1000:[], 2000:[], 3000:[], 4000:[], 5000:[]}
+for row in csv_read:
+    #import ipdb ; ipdb.set_trace()
+    for i, col in enumerate(row):
+        
+        # 0 - <>
+        # 1 - M
+        # 2 - N
+        # 3 - size
+        # 4 - degree min
+        # 5 - degree max
+        # 6 - shortestpath
+        # 7 - 2/3
+        sz = int(row[3])
+        n_rw = int(row[2])
+        avgd_rw = np.mean([int(row[4]), int(row[5])])
+        two3_rw = float(row[7])
+
+        n[sz].append(n_rw)
+        avgd[sz].append(avgd_rw)
+        two3[sz].append(two3_rw)
+
+
+
+fig = plt.figure()
+ax = fig.add_subplot(111, projection='3d')
+
+for i, j in enumerate(n[1000]):
+
+    print(n[1000][i], avgd[1000][i], two3[1000][i])
+
+ax.scatter(n[1000], avgd[1000], two3[1000], c='r', marker='x', label='1000')
+ax.scatter(n[2000], avgd[2000], two3[2000], c='g', marker='+', label='2000')
+ax.scatter(n[3000], avgd[3000], two3[3000], c='b', marker='1', label='3000')
+ax.scatter(n[4000], avgd[4000], two3[4000], c='purple', marker='2', label='4000')
+ax.scatter(n[5000], avgd[5000], two3[5000], c='c', marker='3', label='5000')
+
+#ax.set_xscale('log')
+#ax.set_xticklabels([1,2,3,10,100])
+#ax.set_xticks([1,2,3,10,100])
+
+ax.set_xlabel('N')
+ax.set_ylabel('Average degree')
+ax.set_zlabel('Time to reach 2/3rds of nodes')
+
+ax.legend()
+
+plt.show()

plot_n_avgd_23.py

@@ -0,0 +1,3 @@
+numpy==1.13.3
+matplotlib==2.1.1
+networkx==1.11

requirements.txt

@@ -0,0 +1,15 @@
+#!/bin/bash
+
+for M in {2..5}
+do
+    for N in 1 2 3 10 100
+    do
+        for degree in {5..10}
+        do
+            for size in `seq 1000 1000 5000`
+            do
+                time python sim.py $M $N $size $degree $(( $degree + 10 )) >> log.txt 2>&1
+            done
+        done
+    done
+done