/
generate_data.py
191 lines (157 loc) · 7.81 KB
/
generate_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
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
import numpy as np
import pandas as pd
from random import random, uniform
from sklearn.datasets.samples_generator import make_blobs
import sys
from sklearn import datasets
def generate_spherical_clusters(number_of_samples, number_of_clusters, n_features=2, variances=None, filename=""):
"""
:param number_of_samples: The total number of points equally divided among clusters.
:param number_of_clusters: The number of clusters to generate
:param n_features: The number of features for each sample.
:param variances: The standard deviation of the clusters.
:param filename: The file to store the results
:return:
"""
if variances is None: variances = [0.5 for _ in xrange(number_of_clusters)]
if filename == "":
filename = "./Data/spherical_" + str(number_of_samples) + "_features_" + str(n_features) \
+ "_cluster_" + str(number_of_clusters) + ".csv"
random_state = 170
X, y = make_blobs(n_samples=number_of_samples, centers=number_of_clusters, n_features=n_features,
random_state=random_state, cluster_std=variances)
features = ["features_" + str(i+1) for i in xrange(n_features)]
df = pd.DataFrame()
for i, feature in enumerate(features): df[feature] = X[:, i]
df["class"] = y
df.to_csv(filename, index=False)
return X, y
def _generate_spherical_clusters():
X, y = generate_spherical_clusters(10000, 4)
import matplotlib.pyplot as plt
plt.scatter(X[:, 0], X[:, 1], c=y)
plt.show()
def generate_anisotropically_clusters(number_of_samples, number_of_clusters, n_features=2, variances=None, filename=""):
"""
:param number_of_samples: The total number of points equally divided among clusters.
:param number_of_clusters: The number of clusters to generate
:param n_features: The number of features for each sample.
:param variances: The standard deviation of the clusters.
:param filename: The file to store the results
:return:
"""
if variances is None: variances = [0.5 for _ in xrange(number_of_clusters)]
if filename == "":
filename = "./Data/anisotropically_" + str(number_of_samples) + "_features_" + str(n_features) \
+ "_cluster_" + str(number_of_clusters) + ".csv"
random_state = 170
X, y = make_blobs(n_samples=number_of_samples, centers=number_of_clusters, n_features=n_features,
random_state=random_state, cluster_std=variances)
transformation = np.array([[random() if i == j else uniform(-1, 1) for j in xrange(n_features)] for i in xrange(n_features)])
X = np.dot(X, transformation)
features = ["features_" + str(i + 1) for i in xrange(n_features)]
df = pd.DataFrame()
for i, feature in enumerate(features): df[feature] = X[:, i]
df["class"] = y
df.to_csv(filename, index=False)
return X, y
def _generate_anisotropically_clusters():
X, y = generate_anisotropically_clusters(10000, 4)
import matplotlib.pyplot as plt
plt.scatter(X[:, 0], X[:, 1], c=y)
plt.show()
def generate_varied_clusters(number_of_samples, number_of_clusters, n_features=2, variances=None, filename=""):
"""
:param number_of_samples: The total number of points equally divided among clusters.
:param number_of_clusters: The number of clusters to generate
:param n_features: The number of features for each sample.
:param variances: The standard deviation of the clusters.
:param filename: The file to store the results
:return:
"""
if variances is None: variances = [uniform(1, 3) for _ in xrange(number_of_clusters)]
if filename == "":
filename = "./Data/varied_" + str(number_of_samples) + "_features_" + str(n_features) \
+ "_cluster_" + str(number_of_clusters) + ".csv"
random_state = 170
X, y = make_blobs(n_samples=number_of_samples, centers=number_of_clusters, n_features=n_features,
random_state=random_state, cluster_std=variances)
features = ["features_" + str(i+1) for i in xrange(n_features)]
df = pd.DataFrame()
for i, feature in enumerate(features): df[feature] = X[:, i]
df["class"] = y
df.to_csv(filename, index=False)
return X, y
def _generate_varied_clusters():
X, y = generate_varied_clusters(10000, 4)
import matplotlib.pyplot as plt
plt.scatter(X[:, 0], X[:, 1], c=y)
plt.show()
def points_for_clusters():
no_instances = [100, 1000, 10000, 100000, 1000000]
no_features = [2, 4, 8, 16, 32, 64, 128, 256, 512]
no_clusters = [2, 4, 6, 8]
for no_instance in no_instances:
for no_feature in no_features:
for no_cluster in no_clusters:
print "# ",
sys.stdout.flush()
generate_spherical_clusters(number_of_samples=no_instance, number_of_clusters=no_cluster, n_features=no_feature)
generate_anisotropically_clusters(number_of_samples=no_instance, number_of_clusters=no_cluster, n_features=no_feature)
generate_varied_clusters(number_of_samples=no_instance, number_of_clusters=no_cluster, n_features=no_feature)
print
print
def generate_regression_datasets(n_samples, n_features, noise=25, bias=100):
X = datasets.make_regression(n_samples=n_samples, n_features=n_features, coef=True, noise=noise, bias=bias)
header = ["A" + str(i) for i in xrange(len(X[0][0]))] + ["Dep"]
content = []
for independent, dependent in zip(X[0], X[1]):
content.append(independent.tolist() + [dependent])
filename = "./RData/Regression_" + str(n_samples) + "_" + str(n_features) + "_" + str(noise) + "_" + str(bias) + ".csv"
df = pd.DataFrame(content, columns=header)
df.to_csv(filename, index=False)
def _generate_regression_datasets():
n_samples = [100, 1000, 10000, 100000, 1000000]
no_features = [2, 4, 8, 16, 32, 64, 128, 256, 512]
for n_sample in n_samples:
for no_feature in no_features:
print "# ",
sys.stdout.flush()
generate_regression_datasets(n_samples=n_sample, n_features=no_feature)
print
def generate_classification_dataset(n_samples, n_features, n_informative, weights, n_clusters_per_class=3, n_classes=2):
"""
Constraints:
- Number of informative, redundant and repeated features must sum to less than the number of total features
- n_classes * n_clusters_per_class must be smaller or equal 2 ** n_informative
:param n_samples:
:param n_features:
:param n_informative:
:param weights:
:param n_clusters_per_class:
:param n_classes:
:return:
"""
from sklearn import datasets
X = datasets.make_classification(n_samples=n_samples, n_features=n_features, n_classes=n_classes,
n_informative=n_informative, n_redundant=2, weights=weights)
header = ["A" + str(i) for i in xrange(len(X[0][0]))] + ["Dep"]
content = []
for independent, dependent in zip(X[0], X[1]):
content.append(independent.tolist() + [dependent])
filename = "./CData/Classification_" + str(n_samples) + "_" + str(n_features) + "_" + str(n_classes) + "_" + str(weights[0]*100) + ".csv"
df = pd.DataFrame(content, columns=header)
df.to_csv(filename, index=False)
if __name__ == "__main__":
n_samples = [100, 1000, 10000, 100000, 1000000]
no_features = [8, 16, 32, 64, 128, 256, 512]
weights = [0.1, 0.2, 0.3, 0.4, 0.5]
for n_sample in n_samples:
for no_feature in no_features:
for weight in weights:
print "# ",
sys.stdout.flush()
t_w = [weight, 1-weight]
generate_classification_dataset(n_samples=n_sample, n_features=no_feature, n_informative=2, weights=t_w)
print
print