-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathdata_reader.py
198 lines (166 loc) · 8.5 KB
/
data_reader.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
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
from __future__ import division
import os
import io
import logging
import numpy as np
import data as data
from sklearn.model_selection import train_test_split
class DataReader(object):
def __init__(self, train, dev, test, nb_classes):
self.data = {'train': train, 'dev': dev, 'test': test}
self.nb_classes = nb_classes
self.max_sentence_length = self.get_max_sentence_length()
def get_max_sentence_length(self):
samples = self.data['train']['X'] + self.data['dev']['X'] + \
self.data['test']['X']
max_sentence_length = 0
for sample in samples:
sample_length = len(sample)
if max_sentence_length < sample_length:
max_sentence_length = sample_length
return max_sentence_length
def get_word_embedding(self, path_to_vec,orthonormalized=True):
samples = self.data['train']['X'] + self.data['dev']['X'] + \
self.data['test']['X']
id2word, word2id = data.create_dictionary(samples, threshold=0)
word_vec = data.get_wordvec(path_to_vec, word2id,orthonormalized=orthonormalized)
wvec_dim = len(word_vec[next(iter(word_vec))])
#stores the value of theta for each word
word_complex_phase = data.set_wordphase(word2id)
params = {'word2id':word2id, 'word_vec':word_vec, 'wvec_dim':wvec_dim,'word_complex_phase':word_complex_phase,'id2word':id2word}
return params
def create_batch(self, embedding_params, batch_size = -1):
embed = {'train': {}, 'dev': {}, 'test': {}}
for key in self.data:
embed[key] = {'X':[],'y':[]}
logging.info('Computing embedding for {0}'.format(key))
sorted_data = sorted(zip(self.data[key]['X'],
self.data[key]['y']),
key=lambda z: (len(z[0]), z[1]))
self.data[key]['X'], self.data[key]['y'] = map(list, zip(*sorted_data))
bsize = batch_size
if (batch_size == -1):
bsize = len(self.data[key]['y'])
for ii in range(0, len(self.data[key]['y']), bsize):
batch = self.data[key]['X'][ii:ii + bsize]
embeddings = data.get_index_batch(embedding_params, batch)
# print(embeddings)
embed[key]['X'].append(embeddings)
# print(self.sst_data[key]['y'][ii:ii + batch_size])
embed[key]['y'].append(self.data[key]['y'][ii:ii + bsize])
# sst_embed[key]['X'] = np.vstack(sst_embed[key]['X'])
# print(sst_embed[key]['y'])
embed[key]['y'] = np.array(embed[key]['y'])
# print(sst_embed[key]['y'])
logging.info('Computed {0} embeddings'.format(key))
return embed
class TRECDataReader(DataReader):
def __init__(self, task_dir_path, seed=1111):
self.seed = seed
train = self.loadFile(os.path.join(task_dir_path, 'train_5500.label'))
train, dev = self.train_dev_split(train, train_dev_ratio = 1/9)
test = self.loadFile(os.path.join(task_dir_path, 'TREC_10.label'))
nb_classes = 6
super(TRECDataReader,self).__init__(train, dev, test, nb_classes)
def train_dev_split(self, samples, train_dev_ratio = 1/9):
X_train, X_dev, y_train, y_dev = train_test_split(samples['X'], samples['y'], test_size=train_dev_ratio, random_state=self.seed)
train = {'X': X_train, 'y':y_train}
dev = {'X': X_dev, 'y':y_dev}
return train, dev
def loadFile(self, fpath):
trec_data = {'X': [], 'y': []}
tgt2idx = {'ABBR': 0, 'DESC': 1, 'ENTY': 2,
'HUM': 3, 'LOC': 4, 'NUM': 5}
with io.open(fpath, 'r', encoding='latin-1') as f:
for line in f:
target, sample = line.strip().split(':', 1)
sample = sample.split(' ', 1)[1].split()
assert target in tgt2idx, target
trec_data['X'].append(sample)
trec_data['y'].append(tgt2idx[target])
return trec_data
class SSTDataReader(DataReader):
def __init__(self, task_dir_path, nclasses = 2, seed = 1111):
self.seed = seed
# binary of fine-grained
assert nclasses in [2, 5]
self.nclasses = nclasses
self.task_name = 'Binary' if self.nclasses == 2 else 'Fine-Grained'
train = self.loadFile(os.path.join(task_dir_path, self.task_name,'sentiment-train'))
dev = self.loadFile(os.path.join(task_dir_path, self.task_name, 'sentiment-dev'))
test = self.loadFile(os.path.join(task_dir_path, self.task_name, 'sentiment-test'))
# super().__init__(train, dev, test, nclasses)
super(SSTDataReader,self).__init__(train, dev, test, nclasses)
def loadFile(self, fpath):
sst_data = {'X': [], 'y': []}
with io.open(fpath, 'r', encoding='utf-8') as f:
for line in f:
if self.nclasses == 2:
sample = line.strip().split('\t')
sst_data['y'].append(int(sample[1]))
sst_data['X'].append(sample[0].split())
elif self.nclasses == 5:
sample = line.strip().split(' ', 1)
sst_data['y'].append(int(sample[0]))
sst_data['X'].append(sample[1].split())
assert max(sst_data['y']) == self.nclasses - 1
return sst_data
class BinaryClassificationDataReader(DataReader):
def __init__(self, pos, neg, seed=1111):
self.seed = seed
self.samples, self.labels = pos + neg, [1] * len(pos) + [0] * len(neg)
train, test, dev = self.train_test_dev_split(0.1,1.0/9)
nb_classes = 2
super(BinaryClassificationDataReader,self).__init__(train, test, dev, nb_classes)
def loadFile(self, fpath):
with io.open(fpath, 'r', encoding='latin-1') as f:
return [line.split() for line in f.read().splitlines()]
def train_test_dev_split(self, train_test_ratio = 0.1,train_dev_ratio = 1/9):
X_train, X_test, y_train, y_test = train_test_split(self.samples, self.labels, test_size=train_test_ratio, random_state=self.seed)
X_train, X_dev, y_train, y_dev = train_test_split(X_train, y_train, test_size=train_dev_ratio, random_state=self.seed)
train = {'X': X_train, 'y':y_train}
test = {'X': X_test, 'y':y_test}
dev = {'X': X_dev, 'y':y_dev}
return train, test, dev
class CRDataReader(BinaryClassificationDataReader):
def __init__(self, task_path, seed=1111):
# logging.debug('***** Transfer task : CR *****\n\n')
pos = self.loadFile(os.path.join(task_path, 'custrev.pos'))
neg = self.loadFile(os.path.join(task_path, 'custrev.neg'))
super(CRDataReader,self).__init__(pos, neg, seed)
class MRDataReader(BinaryClassificationDataReader):
def __init__(self, task_path, seed=1111):
# logging.debug('***** Transfer task : MR *****\n\n')
pos = self.loadFile(os.path.join(task_path, 'rt-polarity.pos'))
neg = self.loadFile(os.path.join(task_path, 'rt-polarity.neg'))
super(MRDataReader,self).__init__(pos, neg, seed)
class SUBJDataReader(BinaryClassificationDataReader):
def __init__(self, task_path, seed=1111):
# logging.debug('***** Transfer task : SUBJ *****\n\n')
obj = self.loadFile(os.path.join(task_path, 'subj.objective'))
subj = self.loadFile(os.path.join(task_path, 'subj.subjective'))
super(SUBJDataReader,self).__init__(obj, subj, seed)
class MPQADataReader(BinaryClassificationDataReader):
def __init__(self, task_path, seed=1111):
# logging.debug('***** Transfer task : MPQA *****\n\n')
pos = self.loadFile(os.path.join(task_path, 'mpqa.pos'))
neg = self.loadFile(os.path.join(task_path, 'mpqa.neg'))
super(MPQADataReader,self).__init__(pos, neg, seed)
def data_reader_initialize(reader_type, datasets_dir):
dir_path = os.path.join(datasets_dir, reader_type)
if(reader_type == 'CR'):
return(CRDataReader(dir_path))
if(reader_type == 'MR'):
return(MRDataReader(dir_path))
if(reader_type == 'SUBJ'):
return(SUBJDataReader(dir_path))
if(reader_type == 'MPQA'):
return(MPQADataReader(dir_path))
if(reader_type == 'SST_2'):
dir_path = os.path.join(datasets_dir, 'SST')
return(SSTDataReader(dir_path, nclasses = 2))
if(reader_type == 'SST_5'):
dir_path = os.path.join(datasets_dir, 'SST')
return(SSTDataReader(dir_path, nclasses = 5))
if(reader_type == 'TREC'):
return(TRECDataReader(dir_path))