MovieRec/Movie at master · djhang/MovieRec · GitHub

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
#! /usr/bin/env python
#coding=utf-8
import os
import timeit
import random


predict = open("data-v/predict.txt")

t_user = {}
t_movie= {}
user_dis = {}

def preprocess():
    train = open("data-v/training_set.txt")
    n = 0
    for line in train.readlines():
        terms = line[:-2].split('\t')
        x = float(terms[2])
        if t_user.has_key(terms[0]):
            t_user[terms[0]][terms[1]] = x
        else:
            t_user[terms[0]] = {terms[1]: x}
        if t_movie.has_key(terms[0]):
            t_movie[terms[1]][terms[0]] = x
        else:
            t_movie[terms[1]] = {terms[0] : x}
        n += 1
    #    if n > 1000: break
    print len(t_user)
    return

def manhattn(rating1, rating2):
    n = 0
    sum = 0
    for x in rating1:
        if x in rating2:
            sum += abs(rating1[x] - rating2[x])
            n += 1
    if n == 0:
        return -1
    else:
        return sum/n

def closestUser(uId):
    dis = []
    #print t_user.keys()
    for user in t_user:
        if user != uId:
            d = manhattn(t_user[uId], t_user[user])
            dis.append((d,user))
    dis.sort()
    return dis

def distances():
    for x in t_user:
        dis = closestUser(x)
        user_dis[x] = dis
    return

def input():
    pre = []
    predict = open("data-v/predict.txt")
    for line in predict:
        pre.append(line[:-2])
    return pre

def predict1(pre):
    n = 0
    for i in xrange(0,len(pre)):
        n += 1
        if n % 100 == 0:
            print n
        terms = pre[i].split('\t')
        dis = user_dis[terms[0]]
        count = 0
        sum = 0
        for i in range(0, len(dis)):
            if i > 3:
                break
            if t_user[dis[i][1]].has_key(terms[1]):
                count += 1
                sum += t_user[dis[i][1]][terms[1]]
        if count != 0:
            pre[i] += '\t' + str(sum / count) + '\r\n'
        else:
            pre[i] += '\t' + '4' + '\r\n'
    return pre

def output(ans_predict):
    out = open("output/predict.txt", "w")
    for line in ans_predict:
        out.write(line)
    out.close()


if __name__== "__main__":

    starttime = timeit.default_timer()


    print 'begin'
    preprocess()
    endtime = timeit.default_timer()
    interval=(endtime - starttime)
    print str(interval) + "  seconds"

    distances()
    endtime = timeit.default_timer()
    interval=(endtime - starttime)
    print str(interval) + "  seconds"

    print 'step1'
    pre = input()
    ans_predict = predict1(pre)
    output(ans_predict)

    endtime = timeit.default_timer()
    interval=(endtime - starttime)
    print str(interval) + "  seconds"