Variational_method/predict.py at master · luochuwei/Variational_method · 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
121
122
123
124
125
126
127
128
129
130
131
import numpy as np
import theano
import theano.tensor as T

import cPickle
from collections import OrderedDict
import time

import Variational_Method2

def constuct_batch_data(batch, batch_size, features, x):
    # return 1
    batch_x = x[batch*batch_size:(batch+1)*batch_size]
    max_len_batch_x = max([len(i.split()) for i in batch_x])
    # max_len_batch_x = 70
    # print max_len_batch_x

    for j in xrange(batch_size):
        j_len = len(batch_x[j].split())
        # print j_len
        if j_len < max_len_batch_x:
            for k in xrange(max_len_batch_x - j_len):
                batch_x[j] += ' 0'

    final_x = []
    for jj in xrange(max_len_batch_x):
        batch_word_vec = []
        for kk in xrange(batch_size):
            v = np.zeros(features)
            v[int(batch_x[kk].split()[jj])] = 1
            batch_word_vec.append(v)
        batch_word_vec = np.array(batch_word_vec).transpose()
        final_x.append(batch_word_vec)
    return np.array(final_x).astype('float32')

def trans_to_word(v):
    batch_size = int(v.shape[-1])
    features = int(v.shape[1])

    ans = [[] for k in range(batch_size)]

    for i in range(int(v.shape[0])):
        for j in range(batch_size):
            current_array = v[:,:,j:(j+1)][i].transpose()[0]
            max_item = max(current_array)
            max_item_index = list(current_array).index(max_item)
            ans[j].append(max_item_index)

    return ans
def decode(v, t_steps, z, type_n):
        # x = np.zeros((t_steps+1, self.features))
        # xx = [[] for i in range(t_steps)]
    x = [[] for i in xrange(t_steps+1)]
    x[0] = np.zeros((v.features, z.shape[1]))


    W_Fh = v.params["W_Fh_"+str(type_n)].get_value()
    b_Fh = v.params["b_Fh_"+str(type_n)].get_value()

    W_hhd = v.params["W_hhd_"+str(type_n)].get_value()
    W_xhd = v.params["W_xhd_"+str(type_n)].get_value()

    b_hd = v.params["b_hd_"+str(type_n)].get_value()

    W_hx = v.params["W_hx_"+str(type_n)].get_value()
    b_hx = v.params["b_hx_"+str(type_n)].get_value()

    h = W_Fh.dot(z) + b_Fh

    for t in xrange(t_steps):
        h = np.tanh(W_hhd.dot(h) + W_xhd.dot(x[t]) + b_hd)
        x[t+1] = 1 /(1 + np.exp(-(W_hx.dot(h) + b_hx)))

    return np.array(x[1:])


word_dic = cPickle.load(open(r'E:\Learning\VAE\idea\code\data\word_dic.pkl', 'rb'))
num_word_dic = {}
for i,j in word_dic.iteritems():
    num_word_dic[j] = i

x = cPickle.load(open(r'E:\Learning\VAE\idea\code\data\x_test.pkl','rb'))
fwx = open('test_word.txt','w')
for jj in x:
    for kk in jj.split():
        fwx.write(num_word_dic[int(kk)])
    fwx.write('\n')
fwx.flush()
fwx.close()


Zdim = 200
hidden_units_encoder = 300
hidden_units_decoder = 300
features = 26111
batch_size = 2

print '1 build model'
v = Variational_Method2.VM2(2,hidden_units_encoder,hidden_units_decoder,features,Zdim,0.05,0.001,0.005,0.1,batch_size)
v.load_parameters(r'E:\Learning\VAE\idea\code\para')
print '2 load over'

# x_0batch = constuct_batch_data(0, batch_size, features, x)
print '3 encode'
# f,z = v.encode(x_0batch)
print '4 decode'
# y0 = decode(v,20,f[0],0)

fwww = open(r'r.txt', 'w')
for b in range(len(x)/batch_size):
    x_batch = constuct_batch_data(b, batch_size, features, x)
    f,z = v.encode(x_batch)
    zz = trans_to_word(v.decode_z(int(x_batch.shape[0]), z))
    y0 = trans_to_word(decode(v, 20, f[0], 0))
    y1 = trans_to_word(decode(v, 20, f[1], 1))
    for r in range(2):
        for l in zz[r]:
            if l!=0:
                fwww.write(num_word_dic[l])
        fwww.write('\t')
        for l in y0[r]:
            if l != 0:
                fwww.write(num_word_dic[l])
        fwww.write('\t')
        for l in y1[r]:
            if l !=0:
                fwww.write(num_word_dic[l])
        fwww.write('\n')
fwww.flush()
fwww.close()