

In [3]:

    
import networkx as nx
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import pylab

G = nx.DiGraph()



In [4]:

    
Scores = pd.read_csv("scores.csv")
Scores[:10]

Page Rank



In [5]:

    
G = nx.DiGraph()
for i in range(len(Scores)):
    for x in Scores.columns.values:
        if Scores.ix[i][x] > 0:
            for y in Scores.columns.values:
                if Scores.ix[i][y] > 0:
                    if Scores.ix[i][x] > Scores.ix[i][y]:
                        if (x,y) not in G.edges():
                            G.add_edges_from([(x,y)])



In [6]:

    
pos=nx.spring_layout(G)
nx.draw(G,pos,node_color = 'blue', node_size=2000)
node_labels = {node:node for node in G.nodes()}
nx.draw_networkx_labels(G, pos, labels=node_labels)
pylab.show()



In [7]:

    
nx.pagerank(G)









    Out[7]:





{'3series': 0.07910322957982689,
 '5series': 0.09573750117209036,
 '7series': 0.1216537730762872,
 'A6': 0.05892813273464103,
 'A8': 0.09662922372038388,
 'ES': 0.09559802761764902,
 'LS': 0.1645243436346072,
 'RX': 0.076621128657151,
 'Sclass': 0.12734701517178246,
 'XJ': 0.08385762463558098}



In [8]:

    
sorted(list(nx.pagerank(G)), key=lambda tup: tup[1])









    Out[8]:





['A6', 'A8', 'XJ', 'ES', 'LS', 'RX', 'Sclass', '7series', '5series', '3series']

Weighted Page Rank



In [9]:

    
W = nx.DiGraph()
for i in range(len(Scores)):
    for x in Scores.columns.values:
        if Scores.ix[i][x] > 0:
            for y in Scores.columns.values:
                if Scores.ix[i][y] > 0:
                    if Scores.ix[i][x] > Scores.ix[i][y]:
                        W.add_edges_from([(x,y)])



In [10]:

    
nx.pagerank(W)









    Out[10]:





{'3series': 0.07910322957982689,
 '5series': 0.09573750117209036,
 '7series': 0.1216537730762872,
 'A6': 0.05892813273464103,
 'A8': 0.09662922372038388,
 'ES': 0.09559802761764902,
 'LS': 0.1645243436346072,
 'RX': 0.076621128657151,
 'Sclass': 0.12734701517178246,
 'XJ': 0.08385762463558098}



In [11]:

    
sales = {
    "A6": 20,
    "A8": 12,
    "3series": 220,
    "5series": 60,
    "7series": 14,
    "XJ": 6.6,
    "ES": 135,
    "LS": 30,
    "RX": 120,
    "Sclass": 25
}



In [13]:

    
test = pd.DataFrame({"Sales": sales, "Weighted": nx.pagerank(W), "Unweighted": nx.pagerank(G)})



In [17]:

    
from sklearn.linear_model import LinearRegression



In [29]:

    
model_uw = LinearRegression()
model_uw.fit(test[["Unweighted"]], test.Sales)
print model_uw.coef_
print model_uw.score(test[["Unweighted"]], test.Sales)









    



[-753.4467295]
0.103980708337



In [26]:

    
model_w = LinearRegression()
model_w.fit(test[["Weighted"]], test.Sales)









    Out[26]:





LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)



In [27]:

    
model_w.score(test[["Weighted"]], test.Sales)









    Out[27]:





0.10398070833650408

	ES	LS	RX	A8	A6	3series	5series	7series	XJ	Sclass
0	NaN	NaN	4.0	3.0	NaN	NaN	NaN	5.0	2.0	1.0
1	NaN	NaN	3.0	2.0	2.0	2.0	2.0	2.0	NaN	NaN
2	NaN	4.0	NaN	NaN	NaN	3.0	NaN	3.0	NaN	3.0
3	NaN	2.0	NaN	2.0	NaN	NaN	NaN	2.0	4.0	2.0
4	NaN	3.0	3.0	NaN	NaN	NaN	2.0	NaN	NaN	3.0
5	NaN	NaN	NaN	1.0	NaN	NaN	NaN	3.0	4.0	3.0
6	NaN	2.0	NaN	3.0	NaN	NaN	NaN	4.0	-1.0	3.0
7	NaN	NaN	NaN	3.0	NaN	NaN	NaN	1.0	4.0	2.0
8	NaN	2.0	NaN	4.0	3.0	-2.0	NaN	NaN	NaN	3.0
9	NaN	4.0	NaN	2.0	NaN	NaN	NaN	2.0	NaN	2.0