-
Notifications
You must be signed in to change notification settings - Fork 8
/
data_helpers.py
120 lines (107 loc) · 4.09 KB
/
data_helpers.py
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
import numpy as np
import pandas as pd
import nltk
import re
train_df = pd.read_csv("data/train_text.csv")
test_df = pd.read_csv("data/test_text.csv")
train_sentence_length = max([len(nltk.word_tokenize(x)) for x in train_df['sentence']])
test_sentence_length = max([len(nltk.word_tokenize(x)) for x in test_df['sentence']])
MAX_SENTENCE_LENGTH = max(train_sentence_length, test_sentence_length)
# LABELS_COUNT = 19
LABELS_COUNT = 2
def clean_str(string):
"""
Tokenization/string cleaning for all datasets except for SST.
Original taken from https://github.com/yoonkim/CNN_sentence/blob/master/process_data.py
This is for splitting English, changing all word to lowercase.
"""
string = re.sub(r"[^A-Za-z0-9(),!?\'\`]", " ", string)
string = re.sub(r"\'s", " \'s", string)
string = re.sub(r"\'ve", " \'ve", string)
string = re.sub(r"n\'t", " n\'t", string)
string = re.sub(r"\'re", " \'re", string)
string = re.sub(r"\'d", " \'d", string)
string = re.sub(r"\'ll", " \'ll", string)
string = re.sub(r",", " , ", string)
string = re.sub(r"!", " ! ", string)
string = re.sub(r"\(", " \( ", string)
string = re.sub(r"\)", " \) ", string)
string = re.sub(r"\?", " \? ", string)
string = re.sub(r"\s{2,}", " ", string)
return string.strip().lower()
def load_data_and_labels(path):
# read training data from CSV file
df = pd.read_csv(path)
# Text data
# cause chanyelian data are chinese, and there are no space between sentence and label
# 通过 为 上游 和 中间 贸易商 提供 供应链 金融 服务 解决 其 资金 短缺 问题 0
# => sentence = 通过 为 上游 和 中间 贸易商 提供 供应链 金融 服务 解决 其 资金 短缺 问题
# label = 0
# => 0 => [1 0]
# 1 => [0 1]
#
x_text = df['sentence'].tolist()
sentence = []
label = []
for s in x_text:
sen = s.split()
length = len(sen)
label_temp = sen[-1]
# print(y_temp)
if label_temp == '0':
label_temp = [1, 0]
elif label_temp == '1':
label_temp = [0, 1]
label.append(label_temp)
sentence.append(' '.join(sen[:length-1]))
# Label Data
# y = df['label']
# labels_flat = y.values.ravel()
#
# labels_count = np.unique(labels_flat).shape[0]
#
# # convert class labels from scalars to one-hot vectors
# # 0 => [1 0 0 0 0 ... 0 0 0 0 0]
# # 1 => [0 1 0 0 0 ... 0 0 0 0 0]
# # ...
# # 18 => [0 0 0 0 0 ... 0 0 0 0 1]
# def dense_to_one_hot(labels_dense, num_classes):
# num_labels = labels_dense.shape[0]
# index_offset = np.arange(num_labels) * num_classes
# labels_one_hot = np.zeros((num_labels, num_classes))
# labels_one_hot.flat[index_offset + labels_dense.ravel()] = 1
# return labels_one_hot
#
# labels = dense_to_one_hot(labels_flat, labels_count)
# labels = labels.astype(np.uint8)
# y = df['label'].tolist()
# sentence = np.array(sentence)
# y = np.array(y)
# print(y)
return sentence, label
def batch_iter(data, batch_size, num_epochs, shuffle=True):
"""
Generates a batch iterator for a dataset.
"""
data = np.array(data)
data_size = len(data)
num_batches_per_epoch = int((len(data) - 1) / batch_size) + 1
print("Total {} epochs".format(num_epochs))
print("{} steps for each epoch".format(num_batches_per_epoch))
print("==========")
for epoch in range(num_epochs):
# Shuffle the data at each epoch
print('\033[1;32mepoch {}: \033[0m'.format(epoch))
if shuffle:
shuffle_indices = np.random.permutation(np.arange(data_size))
shuffled_data = data[shuffle_indices]
else:
shuffled_data = data
for batch_num in range(num_batches_per_epoch):
start_index = batch_num * batch_size
end_index = min((batch_num + 1) * batch_size, data_size)
yield shuffled_data[start_index:end_index]
if __name__ == "__main__":
print("Train / Test file created")
#
# load_data_and_labels("data/test_google.csv")