mlp/Compress_test_our2.py at master · koolok/mlp · 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
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
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Tue Dec 12 14:38:10 2017

@author: koolok
"""

from picture2word import picture2word_multi_, reduce_multi
from multiprocessing import Pool
import Mnist
import random
import time
import analyse
import pickle as pk
from sklearn.model_selection import KFold
import numpy as np


def test_vs (images_base,images_labels,word_test,word_test_label) :
    test_set_size = len(word_test)
    # Test uncompress
    pool = Pool()

#    results = pool.starmap_async(picture2word_multi_, zip(images_base,images_labels)).get()
#
#    results = list(zip(*results))

    base = images_base
    base_labels = images_labels

#    results = pool.starmap_async(reduce_multi, zip(word_test,word_test_label)).get()
#
#    pool.close()
#
#    results = list(zip(*results))

    test_set = word_test
    test_set_labels = word_test_label


    miss_classified = 0
    start = time.time()
    for i in range(len(test_set)) :
        word = test_set[i]

        digit = analyse.edit_analyse_multi(word,base,base_labels,k=3)

        if (digit != test_set_labels[i]) :
            miss_classified += 1

    processing_time = (time.time() - start) / test_set_size
    accuracy = ((test_set_size - miss_classified) / test_set_size) * 100

    print("Uncompress : processing time=", processing_time," accuracy=",accuracy)


    # Test compress

#    pool = Pool()
#
#    results = pool.starmap_async(picture2word_multi_, zip(images_base,images_labels)).get()
#
#    results = list(zip(*results))

    base = images_base
    base_labels = images_labels

    results = pool.starmap_async(reduce_multi, zip(word_test,word_test_label)).get()

    pool.close()

    results = list(zip(*results))

    test_set = results[0]
    test_set_labels = results[1]


    miss_classified = 0
    start = time.time()
    for i in range(len(test_set)) :
        word = test_set[i]

        digit = analyse.edit_analyse_multi(word,base,base_labels,k=3)

        if (digit != test_set_labels[i]) :
            miss_classified += 1

    processing_time = (time.time() - start) / test_set_size
    accuracy_cp = ((test_set_size - miss_classified) / test_set_size) * 100

    print("Compress : processing time=", processing_time," accuracy=",accuracy_cp)

    return accuracy,accuracy_cp

def generate_sets (training_set_size=1000, test_set_size=100):
    train_images, train_labels, test_images, test_labels = Mnist.load_mnist()

    # Generate training set
    indices = list(range(len(train_images)))  # Or just range(len(a)) in Python 2
    random.shuffle(indices)

    Images_base = []
    Images_labels = []


    i = 0
    while (i<len(indices)) :
        images_base = []
        images_labels = []

        max_ref = [training_set_size//10] * 10

        i = 0
        while len(images_base) < training_set_size and i < len(indices) :
            if max_ref[train_labels[i]] > 0 :
                images_base.append(train_images[indices[i]])
                images_labels.append(train_labels[indices[i]])
                max_ref[train_labels[i]] -= 1
                indices.pop(i)
            else :
                i += 1

        if (len(images_base)>= training_set_size):
            Images_base.append(images_base)
            Images_labels.append(images_labels)


    # Generate test set
    indices = list(range(len(test_images)))
    random.shuffle(indices)

    Images_test = []
    Images_test_labels= []

    while ( len(indices) >= test_set_size) :
        images_test = []
        images_test_labels= []

        while len(images_test) < test_set_size :
            images_test.append(test_images[indices[0]])
            images_test_labels.append(test_labels[indices[0]])
            indices.pop(0)

        Images_test.append(images_test)
        Images_test_labels.append(images_test_labels)


    return Images_base, Images_labels, Images_test, Images_test_labels


def multi_test (nb_run=100) :
    Images_base = []
    Images_labels = []
    Images_test = []
    Images_test_labels = []

    accuracy = 0
    accuracy_cp = 0

    file_base = open('ourbase.pk', 'rb')
    base = np.array(pk.load(file_base))
    file_base.close()

    file_labels = open('ourlabel.pk', 'rb')
    base_label = np.array(pk.load(file_labels))
    file_labels.close()


    kf = KFold(n_splits=10, shuffle=True)


    for train_index, test_index in kf.split(base):
        Images_base, Images_test = base[train_index], base[test_index]
        Images_labels, Images_test_labels = base_label[train_index], base_label[test_index]

        accuracy_tmp, accuracy_cp_tmp = test_vs(Images_base, Images_labels, Images_test, Images_test_labels)

        nb_run += 1
        accuracy += accuracy_tmp
        accuracy_cp += accuracy_cp_tmp

    accuracy /= 10
    accuracy_cp /= 10

    print("Uncompress : ",accuracy," Compress : ",accuracy_cp)


multi_test()