-
Notifications
You must be signed in to change notification settings - Fork 1
/
data_helpers.py
294 lines (242 loc) · 11.5 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
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
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
# coding=utf-8
from __future__ import absolute_import, division, print_function
import logging
import csv
import re
import random
from random import randrange, shuffle
from sklearn.model_selection import StratifiedShuffleSplit
import pickle
import numpy as np
logging.basicConfig(format = '%(asctime)s - %(levelname)s - %(name)s - %(message)s',
datefmt = '%m/%d/%Y %H:%M:%S',
level = logging.INFO)
logger = logging.getLogger(__name__)
class InputFeatures(object):
"""A single set of features of data."""
def __init__(self, input_ids, input_mask, segment_ids, label_id, weight, target = None):
self.input_ids = input_ids
self.input_mask = input_mask
self.segment_ids = segment_ids
self.label_id = label_id
self.weight = weight
class InputExample(object):
"""A single training/test example for simple sequence classification."""
def __init__(self, guid, text_a, weight, text_b=None, label=None, target = None):
"""Constructs a InputExample.
Args:
guid: Unique id for the example.
text_a: string. The untokenized text of the first sequence. For single
sequence tasks, only this sequence must be specified.
text_b: (Optional) string. The untokenized text of the second sequence.
Only must be specified for sequence pair tasks.
label: (Optional) string. The label of the example. This should be
specified for train and dev examples, but not for test examples.
"""
self.guid = guid
self.text_a = text_a
self.text_b = text_b
self.label = label
self.target = target
self.weight = weight
def _truncate_seq_pair(tokens_a, tokens_b, max_length):
"""Truncates a sequence pair in place to the maximum length."""
# This is a simple heuristic which will always truncate the longer sequence
# one token at a time. This makes more sense than truncating an equal percent
# of tokens from each, since if one sequence is very short then each token
# that's truncated likely contains more information than a longer sequence.
while True:
total_length = len(tokens_a) + len(tokens_b)
if total_length <= max_length:
break
if len(tokens_a) > len(tokens_b):
tokens_a.pop()
else:
tokens_b.pop()
def convert_examples_to_features(examples, label_list, max_seq_length,
tokenizer, output_mode):
"""Loads a data file into a list of `InputBatch`s."""
label_map = {label : i for i, label in enumerate(label_list)}
features = []
for (ex_index, example) in enumerate(examples):
if ex_index % 10000 == 0:
logger.info("Writing example %d of %d" % (ex_index, len(examples)))
tokens_a = tokenizer.tokenize(example.text_a)
tokens_b = None
if example.text_b:
tokens_b = tokenizer.tokenize(example.text_b)
# Modifies `tokens_a` and `tokens_b` in place so that the total
# length is less than the specified length.
# Account for [CLS], [SEP], [SEP] with "- 3"
_truncate_seq_pair(tokens_a, tokens_b, max_seq_length - 3)
else:
# Account for [CLS] and [SEP] with "- 2"
if len(tokens_a) > max_seq_length - 2:
tokens_a = tokens_a[:(max_seq_length - 2)]
# The convention in BERT is:
# (a) For sequence pairs:
# tokens: [CLS] is this jack ##son ##ville ? [SEP] no it is not . [SEP]
# type_ids: 0 0 0 0 0 0 0 0 1 1 1 1 1 1
# (b) For single sequences:
# tokens: [CLS] the dog is hairy . [SEP]
# type_ids: 0 0 0 0 0 0 0
#
# Where "type_ids" are used to indicate whether this is the first
# sequence or the second sequence. The embedding vectors for `type=0` and
# `type=1` were learned during pre-training and are added to the wordpiece
# embedding vector (and position vector). This is not *strictly* necessary
# since the [SEP] token unambiguously separates the sequences, but it makes
# it easier for the model to learn the concept of sequences.
#
# For classification tasks, the first vector (corresponding to [CLS]) is
# used as as the "sentence vector". Note that this only makes sense because
# the entire model is fine-tuned.
tokens = ["[CLS]"] + tokens_a + ["[SEP]"]
segment_ids = [0] * len(tokens)
if tokens_b:
tokens += tokens_b + ["[SEP]"]
segment_ids += [1] * (len(tokens_b) + 1)
input_ids = tokenizer.convert_tokens_to_ids(tokens)
# The mask has 1 for real tokens and 0 for padding tokens. Only real
# tokens are attended to.
input_mask = [1] * len(input_ids)
# Zero-pad up to the sequence length.
padding = [0] * (max_seq_length - len(input_ids))
input_ids += padding
input_mask += padding
segment_ids += padding
assert len(input_ids) == max_seq_length
assert len(input_mask) == max_seq_length
assert len(segment_ids) == max_seq_length
if output_mode == "classification":
label_id = label_map[example.label]
elif output_mode == "regression":
label_id = float(example.label)
else:
raise KeyError(output_mode)
if ex_index < 5:
logger.info("*** Example ***")
logger.info("guid: %s" % (example.guid))
logger.info("tokens: %s" % " ".join(
[str(x) for x in tokens]))
logger.info("input_ids: %s" % " ".join([str(x) for x in input_ids]))
logger.info("input_mask: %s" % " ".join([str(x) for x in input_mask]))
logger.info(
"segment_ids: %s" % " ".join([str(x) for x in segment_ids]))
logger.info("label: %s (id = %d)" % (example.label, label_id))
features.append(
InputFeatures(input_ids=input_ids,
input_mask=input_mask,
segment_ids=segment_ids,
label_id=label_id,
weight = example.weight))
return features
# Load data
def read_examples(input_file, output_mode = 'classification'):
"""Read a list of `InputExample`s from an input file."""
examples = []
labels = []
toxicity = []
weights = []
unique_id = 0
# Comments with the following indentities will have a higher wright in the loss
identity_columns = [
'male', 'female', 'homosexual_gay_or_lesbian', 'christian', 'jewish',
'muslim', 'black', 'white', 'psychiatric_or_mental_illness']
with open(input_file, "r", encoding='utf-8') as reader:
csv_reader = csv.reader(reader, delimiter=',')
for _i, line in enumerate(csv_reader):
if _i == 0:
# Get headers and look for identity columnns
headers = list(line)
# Stores its positions for futher use
interesting_positions = [headers.index(interest_identity) for interest_identity in identity_columns]
else:
# Get toxicity ground truth
target = float(line[1])
if target >= 0.5:
label = "Toxic"
else:
label = "OK"
text = line[2]
text_a = None
text_b = None
m = re.match(r"^(.*) \|\|\| (.*)$", text)
if m is None:
text_a = text
else:
print(text_a, text_b)
text_a = m.group(1)
text_b = m.group(2)
# store class or float toxicity depending on mode
if output_mode != 'classification':
label = target
# Calculate weight
weight = 0.25
# Subgroup:
weight+= 0.25*(sum([float(line[interest])>=0.5 for interest in interesting_positions if line[interest] !=''])>=1)
# Background Positive, Subgroup Negative
weight+=0.25*((target>=0.5)*sum([float(line[interest])<0.5 for interest in interesting_positions if line[interest] !=''])>=1)
# Background Negative, Subgroup Positive
weight+= 0.25*((target<0.5)*sum([float(line[interest])>=0.5 for interest in interesting_positions if line[interest] !=''])>=1)
# Original implementation
'''
# Overall
weights = np.ones((len(x_train),)) / 4
# Subgroup
weights += (train[identity_columns].fillna(0).values>=0.5).sum(axis=1).astype(bool).astype(np.int) / 4
# Background Positive, Subgroup Negative
weights += (( (train['target'].values>=0.5).astype(bool).astype(np.int) +
(train[identity_columns].fillna(0).values<0.5).sum(axis=1).astype(bool).astype(np.int) ) > 1 ).astype(bool).astype(np.int) / 4
# Background Negative, Subgroup Positive
weights += (( (train['target'].values<0.5).astype(bool).astype(np.int) +
(train[identity_columns].fillna(0).values>=0.5).sum(axis=1).astype(bool).astype(np.int) ) > 1 ).astype(bool).astype(np.int) / 4
'''
examples.append(
InputExample(guid=unique_id, text_a=text_a, text_b=text_b, label = label, weight = weight))
labels.append(label)
toxicity.append(target)
unique_id += 1
return examples, labels, toxicity
def read_from_pkl(fname):
# reads pickled dataset, u can pickle the dataset using prepare_dataset
with (open(fname, "rb")) as openfile:
data = pickle.load(openfile)
return data
def read_splits(fname, train_size = 0.7, random_state = 1993):
# Reads pickled data and returns train_test splits, not the fastest but u should jsut run it one time
data = read_from_pkl(fname)
labels = np.array(data['all_label_ids'],dtype= float)
# 5 Toxicity bins
thresholds = [(0,0.2), (0.2,0.4) , (0.4,0.6), (0.6,0.8) ,(0.8,2 )]
idx_train = []
idx_test = []
totals = []
for low,high in thresholds:
# Filter bin
partition = (labels>=low)*(labels<high)
partition_idx = np.where(partition)[0]
# Calculate how many samples we need for each partition
n_train_samples = int(train_size*partition_idx.shape[0])
# Get train indices
#train_part_idx = random.choices(partition_idx, k=n_train_samples)
print(partition_idx.shape[0], " samples in the bin")
partition_idx = list(partition_idx)
shuffle(partition_idx)
train_part_idx = partition_idx[0:n_train_samples]
idx_train+=list(train_part_idx)
print(len(partition_idx), " samples in the bin")
# Get test indices
idx_test += Diff(partition_idx, list(train_part_idx))
print(len(partition_idx), " samples in the bin")
print("Found {} samples. {} train comments and {} val comments".format(labels.shape[0],len(idx_train), len(idx_test)))
return idx_train, idx_test
def Diff(li1, li2):
return list(set(li1) - set(li2))
if __name__ == '__main__':
fname = 'classification_slen84.pkl'
idx_train, idx_test = read_splits(fname, train_size = 0.7, random_state = 1993)
save = {}
save['train'] = idx_train
save['val'] = idx_test
np.save('partition_idx.npy', save)