forked from scikit-adaptation/skada
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathconftest.py
134 lines (114 loc) · 3.18 KB
/
conftest.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
import numpy as np
import pytest
from skada.datasets import DomainAwareDataset, make_shifted_blobs, make_shifted_datasets
collect_ignore_glob = []
# if 'torch' is not installed, we should not attempt
# to run 'collect' for skada/deep modules
try:
import torch # noqa
except ImportError:
collect_ignore_glob.append('skada/deep/*.py')
@pytest.fixture(scope='function', autouse=True)
def set_seed():
np.random.seed(0)
if 'skada/deep/*.py' not in collect_ignore_glob:
torch.manual_seed(0)
@pytest.fixture(scope='session')
def da_reg_dataset():
return make_shifted_datasets(
n_samples_source=20,
n_samples_target=21,
shift="concept_drift",
mean=0.5,
noise=0.3,
label="regression",
random_state=43,
return_dataset=True,
)
@pytest.fixture(scope='session')
def da_reg_datasets():
da_reg_dataset_1 = make_shifted_datasets(
n_samples_source=5,
n_samples_target=10,
shift="concept_drift",
mean=0.5,
noise=0.3,
label="regression",
random_state=42,
return_dataset=True,
)
da_reg_dataset_2 = make_shifted_datasets(
n_samples_source=10,
n_samples_target=5,
shift="concept_drift",
mean=0.5,
noise=0.3,
label="regression",
random_state=42,
return_dataset=True,
)
return da_reg_dataset_1, da_reg_dataset_2
@pytest.fixture(scope='session')
def da_multiclass_dataset():
return make_shifted_datasets(
n_samples_source=20,
n_samples_target=21,
shift="concept_drift",
noise=0.1,
label="multiclass",
random_state=42,
return_dataset=True,
)
@pytest.fixture(scope='session')
def da_binary_dataset():
return make_shifted_datasets(
n_samples_source=20,
n_samples_target=21,
shift="concept_drift",
noise=0.1,
label="binary",
random_state=42,
return_dataset=True,
)
@pytest.fixture(scope='session')
def da_blobs_dataset():
centers = np.array([[0, 0], [1, 1]])
_, n_features = centers.shape
return make_shifted_blobs(
n_samples=100,
centers=centers,
n_features=n_features,
shift=0.13,
random_state=42,
cluster_std=0.05,
return_dataset=True,
)
@pytest.fixture(scope='session')
def da_dataset() -> DomainAwareDataset:
centers = np.array([[0, 0], [1, 1]])
_, n_features = centers.shape
dataset = make_shifted_blobs(
n_samples=100,
centers=centers,
n_features=n_features,
shift=0.13,
random_state=42,
cluster_std=0.05,
return_dataset=True,
)
centers = np.array([[2, 0], [-1, 2]])
_, n_features = centers.shape
dataset2 = make_shifted_blobs(
n_samples=100,
centers=centers,
n_features=n_features,
shift=0.13,
random_state=42,
cluster_std=0.05,
return_dataset=True,
)
return dataset.merge(dataset2, names_mapping={'s': 's2', 't': 't2'})
@pytest.fixture(scope="session")
def tmp_folder(tmpdir_factory):
folder = tmpdir_factory.mktemp("skada_datasets")
return str(folder)