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
|
# Test methods with long descriptive names can omit docstrings
# pylint: disable=missing-docstring
import unittest
import numpy as np
import Orange
from Orange.evaluation.clustering import Silhouette, \
AdjustedMutualInfoScore, ClusteringEvaluation, ClusteringResults
from Orange.clustering.kmeans import KMeans, KMeansModel
class TestClusteringResults(unittest.TestCase):
@staticmethod
def test_init():
data = Orange.data.Table.from_numpy(
None, np.arange(100).reshape((100, 1)))
res = ClusteringResults(data=data, nmethods=2, nrows=100)
res.actual[:50] = 0
res.actual[50:] = 1
res.predicted = np.vstack((res.actual, res.actual))
expected = [1.0, 1.0]
np.testing.assert_almost_equal(AdjustedMutualInfoScore(res), expected)
class TestClusteringEvaluation(unittest.TestCase):
def test_init(self):
res = ClusteringEvaluation(k=42)
self.assertEqual(res.k, 42)
def test_kmeans(self):
table = Orange.data.Table('iris')
cr = ClusteringEvaluation(k=3)(table, learners=[KMeans(n_clusters=2),
KMeans(n_clusters=3),
KMeans(n_clusters=5)])
expected = [0.68081362, 0.55259194, 0.48851755]
np.testing.assert_almost_equal(Silhouette(cr), expected, decimal=2)
expected = [0.65383807, 0.75511917, 0.68721092]
np.testing.assert_almost_equal(AdjustedMutualInfoScore(cr),
expected, decimal=2)
self.assertIsNone(cr.models)
cr = ClusteringEvaluation(k=3, store_models=True)(
table, learners=[KMeans(n_clusters=2)])
self.assertEqual(cr.models.shape, (3, 1))
self.assertTrue(all(isinstance(m, KMeansModel)
for m in cr.models.flatten()))
|
