-
Notifications
You must be signed in to change notification settings - Fork 0
/
hybrid.py
217 lines (171 loc) · 6.44 KB
/
hybrid.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from sklearn import svm
import networkx as nx
import json
import math
number_of_thresholds = 11
genre = dict()
genre['Action']=0
genre['Adventure']=1
genre['Animation']=2
genre['Children\'s']=3
genre['Comedy']=4
genre['Crime']=5
genre['Documentary']=6
genre['Drama']=7
genre['Fantasy']=8
genre['Film-Noir']=9
genre['Horror']=10
genre['Musical']=11
genre['Mystery']=12
genre['Romance']=13
genre['Sci-Fi']=14
genre['Thriller']=15
genre['War']=16
genre['Western']=17
def getGenreVector(genreString):
vec = [0]*len(genre)
for s in genreString.split('|'):
vec[genre[s]]=1
return vec
def getThreshold(t):
return (t*1.0/(number_of_thresholds-1))
def parse(path):
f = open(path)
for line in f:
yield json.loads(line)
def getDF(path):
i = 0
df = {}
for d in parse(path):
df[i] = d
i += 1
return pd.DataFrame.from_dict(df, orient='index')
def calcuate_similarity(pivot_table, user_data, product_data, i, j):
if i==j:
return 0
common = (pivot_table[i]*pivot_table[j]).nonzero()
val=0
for k in common[0]:
diff= ( pivot_table[i][k]-user_data.iloc[i, 0] )*( pivot_table[j][k]-user_data.iloc[j, 0] )
val = val+diff
return val/ ( max(user_data.iloc[i, 1], 1)*max(user_data.iloc[j, 1], 1) )
#create pandas dataframe
df = pd.read_csv('dataset/ml-1m/ratings.dat', header=None, names=['reviewerID', 'movieID', 'overall', 'unixReviewTime'], sep=':+', engine='python')
df.sort_values('unixReviewTime')
movieDF = pd.read_csv('dataset/ml-1m/movies.dat', header=None, names=['movieID', 'name', 'genre'], sep='::', engine='python')
#create product data
product_data = pd.DataFrame(df.groupby('movieID')['overall'].agg([np.mean, np.std, 'count'])).fillna(1)
top_product = product_data.sort_values('count').tail(100).index.values
df = df[df.movieID.isin(top_product)]
print len(df)
split_time = df['unixReviewTime'].quantile([.75])[0.75]
after = df[df.unixReviewTime>split_time]
before = df[df.unixReviewTime<=split_time]
user_before = before.reviewerID.unique().tolist()
user_after = after.reviewerID.unique().tolist()
common_users = set(user_before).intersection(set(user_after))
before = before[before.reviewerID.isin(common_users)]
#create user data
user_data = pd.DataFrame(before.groupby('reviewerID')['overall'].agg([np.mean, np.std, 'count'])).fillna(1)
print "user_data"
print user_data.describe();
user_data = user_data[user_data['count'] > 4]
accepted_users = user_data.index.values
print len(accepted_users)
before = before[before.reviewerID.isin(accepted_users)]
pivoted_after = after.pivot(index='reviewerID', columns='movieID', values='overall').fillna(0)
#convert before dataframe to numpy array
numpy_array = before.as_matrix(['reviewerID', 'movieID', 'overall'])
#create product purchase matrix
rows, row_pos = np.unique(numpy_array[:, 0], return_inverse=True)
cols, col_pos = np.unique(numpy_array[:, 1], return_inverse=True)
pivot_table = np.zeros((len(rows), len(cols)), dtype=numpy_array.dtype)
pivot_table[row_pos, col_pos] = numpy_array[:, 2]
dense_table = pivot_table
print product_data.describe();
print len(product_data);
number_of_sim_users = 5
result_precision = np.zeros((number_of_thresholds), dtype=np.float64);
result_recall = np.zeros((number_of_thresholds), dtype=np.float64);
result_f_score = np.zeros((number_of_thresholds), dtype=np.float64);
for target in range(len(accepted_users)):
X = []
Y = []
for x in np.nonzero(pivot_table[target])[0]:
# print movieDF[movieDF.movieID==cols[x]].iloc[0, 2]
X.append(getGenreVector(movieDF[movieDF.movieID==cols[x]].iloc[0, 2]))
Y.append(pivot_table[target][x])
classes = len(set(Y))
if(classes>1):
clf = svm.SVC()
clf.fit(X, Y)
for x in np.where(pivot_table[target]==0)[0]:
vec = getGenreVector(movieDF[movieDF.movieID==cols[x]].iloc[0, 2])
prediction = clf.predict([vec])
dense_table[target][x] = prediction[0]
else:
for x in np.where(pivot_table[target]==0)[0]:
dense_table[target][x] = Y[0]
print dense_table
for target in range(len(accepted_users)):
print "target :", target
sim = np.zeros((len(accepted_users)), dtype=np.float64)
for x in range(len(accepted_users)):
sim[x] = calcuate_similarity(dense_table, user_data, product_data, target, x)
sim_users = np.argpartition(sim, -number_of_sim_users)[-number_of_sim_users:]
purchase_record = set([])
for u in sim_users:
for x in np.nonzero(pivot_table[u])[0]:
purchase_record.add(x)
purchase_record = list(purchase_record)
recommend_prob = np.zeros((len(purchase_record)), dtype=np.float64)
weight_sum = np.zeros((len(purchase_record)), dtype=np.float64)
for i in range(len(purchase_record)):
for u in sim_users:
if(dense_table[u][purchase_record[i]]!=0):
recommend_prob[i] += sim[u]*(dense_table[u][purchase_record[i]]-user_data.iloc[u, 0])
weight_sum[i] += sim[u]
for i in range(len(purchase_record)):
recommend_prob[i] = user_data.iloc[target, 0] + recommend_prob[i]/weight_sum[i]
min_rating = recommend_prob.min()
max_rating = recommend_prob.max()
recommend_prob = np.subtract(recommend_prob, min_rating)
recommend_prob = np.true_divide(recommend_prob, (max_rating - min_rating))
# print recommend_prob.mean(), recommend_prob.min(), recommend_prob.max()
for t in range(number_of_thresholds):
threshold = getThreshold(t)
recommendation_list = np.where(recommend_prob>threshold)[0]
count=0
after_purchased_count = len(after[after.reviewerID==rows[target]])
for x in recommendation_list:
if(len(after[(after.reviewerID==rows[target]) & (after.movieID==cols[purchase_record[x]])])>0):
count+=1
print t, count, len(recommendation_list)
if len(recommendation_list)>0 :
precision = count*1.0/len(recommendation_list)
result_precision[t] += precision
recall = count*1.0/after_purchased_count
result_recall[t] += recall
if recall>1:
print "recall"
if((precision+recall)>0):
f = 2.0*precision*recall/(precision+recall)
result_f_score[t] += f
# print "count :", count
# print "precision :", precision
if target>=50:
break;
result_precision = np.true_divide(result_precision, 51)
result_recall = np.true_divide(result_recall, 51)
result_f_score = np.true_divide(result_recall, 51)
f = open('hybrid.csv', 'w+')
for i in range(number_of_thresholds):
s = str(getThreshold(i))+", "+str(result_precision[i])+", "+str(result_recall[i])+", "+str(result_f_score[i])+"\n"
f.write(s)
f.close()
plt.plot([getThreshold(i) for i in range(number_of_thresholds)], result_precision)
plt.axis([0.0, 1.0, 0.0, 1.0])
plt.show()