-
Notifications
You must be signed in to change notification settings - Fork 1
/
create_data.py
59 lines (54 loc) · 2.07 KB
/
create_data.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
import csv
import random
import numpy as np
def dataset_count(T, training_percent):
training_count = (int)(training_percent * T)
validation_count = T - training_count
return training_count, validation_count
def reservoir_sampling(indices, training_count, validation_count):
training_samples = random.sample(indices, training_count)
validation_samples = [index for index in indices if index not in training_samples]
return training_samples, validation_samples
def get_label(label):
label = label.strip()
if label == 'N':
return '0'
elif label == 'A':
return '1'
elif label == 'O':
return '2'
else:
return '3'
def create_dataset(samples, labels, test_labels, training_samples, validation_samples):
writer = open('training_set.csv', 'w')
for index in training_samples:
writer.write(samples[index] + '\n')
writer.close()
writer = open('training_labels.txt', 'w')
for index in training_samples:
writer.write(get_label(labels[index]) + '\n')
writer.close()
writer = open('validation_set.csv', 'w')
for index in validation_samples:
writer.write(samples[index] + '\n')
writer.close()
writer = open('validation_labels.txt', 'w')
for index in validation_samples:
writer.write(get_label(labels[index]) + '\n')
writer.close()
writer = open('test_labels.txt', 'w')
for label in test_labels:
writer.write(get_label(label) + '\n')
writer.close()
if __name__ == "__main__":
training_count, validation_count = dataset_count(7000, 0.8)
samples = []
reader = csv.reader(open('features.csv', 'r'))
for row in reader:
samples.append(','.join(row))
training_samples, validation_samples = reservoir_sampling(range(7000), training_count, validation_count)
with open('training/reference_train_test.txt') as f:
training_labels = f.readlines()
with open('testing/reference_test.txt') as f:
test_labels = f.readlines()
create_dataset(samples, training_labels, test_labels, training_samples, validation_samples)