-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathsynthetic_data_generation.py
146 lines (132 loc) · 4.92 KB
/
synthetic_data_generation.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
#!/usr/bin/env python
# coding: utf-8
import pickle
import argparse
import numpy as np
import pandas as pd
import stan_base_class as stanb
import utils
def synthetic_dataset_3d(random_seed, size=200):
'''
Args:
random_seed (int):
size (int): number of points generated per distribution
Returns:
df (DataFrame): explanatory variables and response variable (target)
target (str): response variable name
'''
# set random seed
np.random.seed(random_seed)
# distribution
list_mu = [(0, -5, -5), (5, 0, -5), (5, 5, 0)]
list_cov = [([5, 0, 0], [0, 1, 0], [0, 0, 1]),
([1, 0, 0], [0, 5, 0], [0, 0, 1]),
([1, 0, 0], [0, 1, 0], [0, 0, 5])]
# Initialize
target = 'target'
np_x = np.array([])
np_y = np.array([])
np_z = np.array([])
# generation
for mu, cov in zip(list_mu, list_cov):
values = np.random.multivariate_normal(mu, cov, size)
np_x = np.hstack([np_x, values[:, 0]])
np_y = np.hstack([np_y, values[:, 1]])
np_z = np.hstack([np_z, values[:, 2]])
df_dataset = pd.DataFrame(np.dstack([np_x, np_y, np_z])[0], columns=['x1', 'x2', 'x3'])
np.random.normal(loc=0, scale=2, size=np_x.size)
df_dataset[target] = df_dataset.sum(axis=1) + np.random.normal(loc=0, scale=2, size=np_x.size)
return df_dataset, target
def synthetic_dataset_5d(random_seed, size=200):
'''
Args:
random_seed (int):
size (int): number of points generated per distribution
Returns:
df_dataset (DataFrame): explanatory variables and response variable (target)
target (str): response variable name
'''
# set random seed
np.random.seed(random_seed)
# distribution
list_mu = [(0, -5, -5, -5, -5),
(5, 0, -5, -5, -5),
(5, 5, 0, -5, -5),
(5, 5, 5, 0, -5),
(5, 5, 5, 5, 0)]
list_cov = np.zeros((5, 5, 5))
for c in range(5):
for i in range(5):
for j in range(5):
if i==j:
if i==c:
list_cov[c, i, j] = 5
else:
list_cov[c, i, j] = 1
elif i < j:
list_cov[c, i, j] = np.random.randn()
elif i > j:
list_cov[c, i, j] = list_cov[c, j, i]
else:
print('[ERROR]')
# Initialize
target = 'target'
np_x1 = np.array([])
np_x2 = np.array([])
np_x3 = np.array([])
np_x4 = np.array([])
np_x5 = np.array([])
# generation
for mu, cov in zip(list_mu, list_cov):
values = np.random.multivariate_normal(mu, cov, size)
np_x1 = np.hstack([np_x1, values[:, 0]])
np_x2 = np.hstack([np_x2, values[:, 1]])
np_x3 = np.hstack([np_x3, values[:, 2]])
np_x4 = np.hstack([np_x4, values[:, 3]])
np_x5 = np.hstack([np_x5, values[:, 4]])
df_dataset = pd.DataFrame(np.dstack([np_x1, np_x2, np_x3, np_x4, np_x5])[0], columns=['x1', 'x2', 'x3', 'x4', 'x5'])
np.random.normal(loc=0, scale=2, size=np_x1.size)
df_dataset[target] = df_dataset.sum(axis=1) + np.random.normal(loc=0, scale=2, size=np_x1.size)
return df_dataset, target
def get_parser():
parser = argparse.ArgumentParser(description='help')
parser.add_argument('-d', '--dataset', default='5d', choices=['3d', '5d'],
help='types of generated dataset')
parser.add_argument('-s', '--save_dir', default='example',
help='save directory name')
parser.add_argument('-r', '--random_seed', default=0,
help='random seed for dataset generation')
return parser.parse_args()
def main():
args = get_parser()
print(f'[INFO] Save directory is {args.save_dir}')
# make directories
utils.make_dir(args.save_dir)
data_dir = args.save_dir + '/data/'
utils.make_dir(data_dir)
# dataset generation
print(f'[INFO] Start generating {args.dataset} dataset')
if args.dataset=='3d':
df_dataset, target = synthetic_dataset_3d(args.random_seed)
elif args.dataset=='5d':
df_dataset, target = synthetic_dataset_5d(args.random_seed)
else:
raise NotImplementedError()
# generate variable information
df_var = pd.DataFrame(columns=['item_name_other', 'item_type'])
df_var['item_name_other'] = df_dataset.columns
df_var['item_type'] = 'continuous'
print('[INFO] Done')
# display(df_dataset)
# Save
print('[INFO] Save dataset')
df_X = df_dataset.drop(target, axis=1)
df_y = pd.DataFrame(df_dataset[target])
pickle.dump(df_X, open(data_dir + 'df_X.pkl','wb'))
pickle.dump(df_y, open(data_dir + 'df_y.pkl','wb'))
df_X.to_csv(data_dir+'df_X.csv')
df_y.to_csv(data_dir+'df_y.csv')
df_var.to_csv(data_dir+'df_var.csv')
print('[INFO] Done')
if __name__ == "__main__":
main()