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
|
from umap import UMAP
import pytest
try:
# works for sklearn>=0.22
from sklearn.manifold import trustworthiness
except ImportError:
# this is to comply with requirements (scikit-learn>=0.20)
# More recent versions of sklearn have exposed trustworthiness
# in top level module API
# see: https://github.com/scikit-learn/scikit-learn/pull/15337
from sklearn.manifold.t_sne import trustworthiness
def test_densmap_trustworthiness(nn_data):
data = nn_data[:50]
embedding, rad_h, rad_l = UMAP(
n_neighbors=10,
min_dist=0.01,
random_state=42,
n_epochs=100,
densmap=True,
output_dens=True,
).fit_transform(data)
trust = trustworthiness(data, embedding, n_neighbors=10)
assert (
trust >= 0.72
), "Insufficiently trustworthy embedding for" "nn dataset: {}".format(trust)
@pytest.mark.skip()
def test_densmap_trustworthiness_random_init(nn_data): # pragma: no cover
data = nn_data[:50]
embedding = UMAP(
n_neighbors=10,
min_dist=0.01,
random_state=42,
init="random",
densmap=True,
).fit_transform(data)
trust = trustworthiness(data, embedding, n_neighbors=10)
assert (
trust >= 0.75
), "Insufficiently trustworthy embedding for" "nn dataset: {}".format(trust)
def test_densmap_trustworthiness_on_iris(iris):
densmap_iris_model = UMAP(
n_neighbors=10,
min_dist=0.01,
random_state=42,
densmap=True,
verbose=True,
).fit(iris.data)
embedding = densmap_iris_model.embedding_
trust = trustworthiness(iris.data, embedding, n_neighbors=10)
assert (
trust >= 0.97
), "Insufficiently trustworthy embedding for" "iris dataset: {}".format(trust)
with pytest.raises(NotImplementedError):
densmap_iris_model.transform(iris.data[:10])
with pytest.raises(ValueError):
densmap_iris_model.inverse_transform(embedding[:10])
def test_densmap_trustworthiness_on_iris_supervised(iris):
densmap_iris_model = UMAP(
n_neighbors=10,
min_dist=0.01,
random_state=42,
densmap=True,
verbose=True,
).fit(iris.data, y=iris.target)
embedding = densmap_iris_model.embedding_
trust = trustworthiness(iris.data, embedding, n_neighbors=10)
assert (
trust >= 0.95
), "Insufficiently trustworthy embedding for" "iris dataset: {}".format(trust)
|
