firstStepWithTensorflow/embedding.py at master · tuchao1996/firstStepWithTensorflow · 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
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
#!/usr/bin/python
# -*- coding: utf-8 -*-

# 对影评数据进行情感分析的神经网络
# 训练一个情感分析模型，以预测某条评价总体是好评（1），还是差评（0）
# 字符串值term转换为特征矢量
# 嵌套理论知识：
# https://developers.google.cn/machine-learning/crash-course/embeddings/obtaining-embeddings

import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

# 设置
import collections
import math
import time

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import tensorflow as tf
from IPython import display
from sklearn import metrics

# tf.keras 包含文件下载和缓冲工具，用来检索数据集
tf.logging.set_verbosity(tf.logging.ERROR)
train_url = 'https://storage.googleapis.com/mledu-datasets/sparse-data-embedding/train.tfrecord'
train_path = tf.keras.utils.get_file(train_url.split('/')[-1], train_url)
test_url = 'https://storage.googleapis.com/mledu-datasets/sparse-data-embedding/test.tfrecord'
test_path = tf.keras.utils.get_file(test_url.split('/')[-1], test_url)

# 构建输入管道
def _parse_function( record ):
    '''Extract features and labels
    '''
    features = {
        'terms' : tf.VarLenFeature( dtype=tf.string ),
        'labels' : tf.FixedLenFeature( shape=[1], dtype=tf.float32 )
    }
    parsed_features = tf.parse_single_example( record, features )
    terms = parsed_features['terms'].values
    labels = parsed_features['labels']
    return {'terms':terms}, labels

def _input_fn( input_filenames, num_epochs=None, shuffle=True ):
    ds = tf.data.TFRecordDataset( input_filenames )
    ds = ds.map( _parse_function )

    if shuffle:
        ds = ds.shuffle( 10000 )

    ds = ds.padded_batch( 25, ds.output_shapes )

    ds = ds.repeat( num_epochs )

    features, labels = ds.make_one_shot_iterator().get_next()

    return features, labels

def modelDNN( feature_columns, train_path, test_path ):
    my_optimizer = tf.train.AdagradDAOptimizer( learning_rate=0.1, global_step=np.int64(1000) )
    my_optimizer = tf.contrib.estimator.clip_gradients_by_norm( my_optimizer, 5.0 )

    classifier = tf.estimator.DNNClassifier(
        feature_columns=feature_columns,
        hidden_units=[ 10, 10 ],
        optimizer=my_optimizer
    )
    classifier.train(
        input_fn=lambda:_input_fn([train_path]),
        steps=1000
    )

    evaluation_metrics = classifier.evaluate(
        input_fn=lambda:_input_fn([train_path]),
        steps=1000
    )
    print( 'Training set metrics:' )
    for m in evaluation_metrics:
        print( m, evaluation_metrics[m] )
    print( '---'*20 )

    evaluation_metrics = classifier.evaluate( input_fn=lambda:_input_fn([test_path]), steps=1000 )
    print( 'Test set metrics:' )
    for m in evaluation_metrics:
        print( m, evaluation_metrics[m] )
    print( '---'*20 )

def trainEmbeddingDNN( train_path, test_path ):
    f = open( 'C:/Users/tuchao1996/Desktop/terms.txt', 'r', encoding='utf8' )
    informative_terms = list( set(f.read().split()) )

    terms_feature_column = tf.feature_column.categorical_column_with_vocabulary_list(key="terms", vocabulary_list=informative_terms)
    terms_embedding_column = tf.feature_column.embedding_column( terms_feature_column, dimension=2 )
    feature_columns = [terms_embedding_column]

    modelDNN( feature_columns, train_path, test_path )

if __name__ == '__main__' :
    start_time = time.time()
    print( time.strftime('%Y-%m-%d %X')+' Program has been launched!' )
    print( '---'*30 )

    # informative_terms = ("bad", "great", "best", "worst", "fun", "beautiful",
    #                     "excellent", "poor", "boring", "awful", "terrible",
    #                     "definitely", "perfect", "liked", "worse", "waste",
    #                     "entertaining", "loved", "unfortunately", "amazing",
    #                     "enjoyed", "favorite", "horrible", "brilliant", "highly",
    #                     "simple", "annoying", "today", "hilarious", "enjoyable",
    #                     "dull", "fantastic", "poorly", "fails", "disappointing",
    #                     "disappointment", "not", "him", "her", "good", "time",
    #                     "?", ".", "!", "movie", "film", "action", "comedy",
    #                     "drama", "family", "man", "woman", "boy", "girl")

    # 任务六：改进参数
    # informative_terms = None
    # f = open( 'C:/Users/tuchao1996/Desktop/terms.txt', 'r', encoding='utf8' )
    # informative_terms = list( set(f.read().split()) )

    # terms_feature_column = tf.feature_column.categorical_column_with_vocabulary_list(key="terms", vocabulary_list=informative_terms)

    # 优化器
    # my_optimizer = tf.train.Ada gradDAOptimizer( learning_rate=0.1, global_step=np.int64(1000) )
    # my_optimizer = tf.contrib.estimator.clip_gradients_by_norm( my_optimizer, 5.0 )

    # # 任务一：使用具有稀疏输入和显示词汇表的线性模型
    # feature_columns = [ terms_feature_column ]
    # classifier = tf.estimator.LinearClassifier(
    #     feature_columns=feature_columns,
    #     optimizer=my_optimizer
    # )
    # classifier.train( input_fn=lambda:_input_fn([train_path]), steps=1000 )
    # evaluation_metrics = classifier.evaluate( input_fn=lambda:_input_fn([train_path]), steps=1000 )

    # print( 'Training set metrics:' )
    # for m in evaluation_metrics:
    #     print( m, evaluation_metrics[m] )
    # print( '---'*20 )

    # evaluation_metrics = classifier.evaluate( input_fn=lambda:_input_fn([test_path]), steps=1000 )

    # print( 'Test set metrics:' )
    # for m in evaluation_metrics:
    #     print( m, evaluation_metrics[m] )
    # print( '---'*20 )

    # 任务二：使用深度神经网络（DNN）模型
    # feature_columns = [tf.feature_column.indicator_column(terms_feature_column)]
    # classifier = tf.estimator.DNNClassifier(
    #     feature_columns=feature_columns,
    #     hidden_units=[ 20, 20 ],
    #     optimizer=my_optimizer
    # )
    # try:
    #     classifier.train( input_fn=lambda:_input_fn([train_path]), steps=1000 )
    #     evaluation_metrics = classifier.evaluate( input_fn=lambda:_input_fn([train_path]), steps=1 )
    #     print( 'Training set metrics:' )
    #     for m in evaluation_metrics:
    #         print( m, evaluation_metrics[m] )
    #     print( '---'*20 )

    #     evaluation_metrics = classifier.evaluate( input_fn=lambda:_input_fn([test_path]), steps=1 )

    #     print( 'Test set metrics:' )
    #     for m in evaluation_metrics:
    #         print( m, evaluation_metrics[m] )
    #     print( '---'*20 )

    # except ValueError as err :
    #     print( err )

    # 任务三：在DNN中使用嵌入
    # 嵌入列会将稀疏数据作为输入，返回一个低纬度密集矢量作为输出
    # terms_embedding_column = tf.feature_column.embedding_column( terms_feature_column, dimension=2 )
    # feature_columns = [terms_embedding_column]

    # classifier = tf.estimator.DNNClassifier(
    #     feature_columns=feature_columns,
    #     hidden_units=[ 10, 10 ],
    #     optimizer=my_optimizer
    # )

    # try:
    #     classifier.train( input_fn=lambda:_input_fn([train_path]), steps=1000 )
    #     evaluation_metrics = classifier.evaluate( input_fn=lambda:_input_fn([train_path]), steps=1000 )
    #     print( 'Training set metrics:' )
    #     for m in evaluation_metrics:
    #         print( m, evaluation_metrics[m] )
    #     print( '---'*20 )

    #     evaluation_metrics = classifier.evaluate( input_fn=lambda:_input_fn([test_path]), steps=1000 )
    #     print( 'Test set metrics:' )
    #     for m in evaluation_metrics:
    #         print( m, evaluation_metrics[m] )
    #     print( '---'*20 )

    # except ValueError as err :
    #     print( err )

    # 任务四：确信模型中存在嵌入
    # for i in range( len(classifier.get_variable_names()) ):
    #     print( classifier.get_variable_names()[i] )
    # print( classifier.get_variable_value( 'dnn/input_from_feature_columns/input_layer/terms_embedding/embedding_weights' ).shape )

    # # 任务五：检查嵌入
    # embedding_matrix = classifier.get_variable_value( 'dnn/input_from_feature_columns/input_layer/terms_embedding/embedding_weights' )

    # for term_index in range( len(informative_terms) ):
    #     term_vector = np.zeros( len(informative_terms) )
    #     term_vector[term_index] = 1
    #     embedding_xy = np.matmul( term_vector, embedding_matrix )
    #     plt.text( embedding_xy[0], embedding_xy[1], informative_terms[term_index] )

    # plt.hold()
    # plt.plot( [-1.2*embedding_matrix.min(),1.2*embedding_matrix.max()], [0,0] )

    # plt.rcParams['figure.figsize'] = (12, 12)
    # plt.xlim( 1.2*embedding_matrix.min(), 1.2*embedding_matrix.max() )
    # plt.ylim( 1.2*embedding_matrix.min(), 1.2*embedding_matrix.max() )
    # plt.show()

    trainEmbeddingDNN(train_path, test_path)

    print( time.strftime('%Y-%m-%d %X')+' Program has been terminated!' )
    stop_time = time.time()
    spendTime = str( round( stop_time-start_time, 2 ) )
    print( 'Running time -> ' + spendTime + ' seconds.' )