Netflix_Recommender/SparseMatrix.py at master · csjungminlim/Netflix_Recommender · 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
# -*- coding: utf-8 -*-

from scipy import sparse
from scipy.sparse import csr_matrix
from scipy.stats import uniform
import numpy as np
import SVD
from SGD import MF
from Find_Neighbors import Find_Neighbors

def construct_sparse_matrix():

    #/Users/limjungmin/Netflix_Recommender/netflix-prize-data/combined_data_1~4.txt
    data_file = open("/Users/limjungmin/Netflix_Recommender/netflix-prize-data/combined_data_1.txt")
    movie_row = []
    movie_column = []
    data = []
    already_used = []
    user_index = {}
    col_index = 0
    printcounter = 0
    check_example = 0

    for line in data_file:
        if check_example == 1:
            break
        if not line.strip():                            # 빈 줄 건너뛰기
            continue
        if ":" in line:                                 # movie 구분점
            printcounter += 1
            print printcounter
            length = line.__len__() - 2
            movieID = line[:length].rstrip()            # "1:" 경우 1, "234:" 경우 234
            movie_index = eval(movieID) - 1             # indexing 처리
            if eval(movieID) == 20:                     # 20번째 영화에서 정지(testing 때문)
                check_example = 1

        else:
            userID = line.partition(",")
            rating = userID[2].partition(",")[0]
            if userID[0] not in user_index.keys():      # 사용자 ID 중복이 아니라면,
                user_index[userID[0]] = col_index
                movie_row.append(movie_index)
                movie_column.append(col_index)
                data.append(eval(rating))
                col_index += 1

            else:
                x = user_index[userID[0]]               # 기존에 존재하는 사용자라면,
                movie_row.append(movie_index)
                movie_column.append(x)
                data.append(eval(rating))

    # list -> ndarray로 변환
    a = np.asarray(movie_row)
    b = np.asarray(movie_column)
    c = np.asarray(data)
    c = c.astype('float')

#    SVD.sgd(a, b, c)
    matrix_data = sparse.coo_matrix((c, (a, b)))        # coo_matrix((data, (i,j))) 형태의 매트릭스로 변환
    np.set_printoptions(threshold=np.inf)
    print matrix_data.toarray()

    '''

    matrix_data = sparse.csr_matrix(matrix_data)
    R = matrix_data.toarray()
    mf = MF(R, K=2, alpha=0.02, beta=0.001, iterations=120)
    mf.train()
    print mf.full_matrix()
    neighbor_module = Find_Neighbors(matrix_data)
    Find_Neighbors._compare_user_(neighbor_module)
    movie_aspect_vector = mf.get_movie_SVD()
    item_aspect_vector = mf.get_item_SVD()

    matrix_data = sparse.coo_matrix((c, (a, b)))
    matrix_data = sparse.csr_matrix(matrix_data)
    R = matrix_data
    print "done"

    '''

if __name__ == '__main__':
    sparse_matrix = construct_sparse_matrix()