def test_ordering_on(self, data):
def p(val, obj=None):
self.assert_(val >= 0 and val <= 100)
self.assertIsInstance(val, float)
matrix = instance_distance_matrix(data, Euclidean(), progress_callback=p)
root1 = HierarchicalClustering(matrix, progress_callback=p)
root2 = hier.clone(root1)
order_leaves_py(root1, matrix, progress_callback=p)
order_leaves_cpp(root2, matrix, progress_callback=p)
def score(mapping):
sum = 0.0
for i in range(matrix.dim - 1):
sum += matrix[mapping[i], mapping[i + 1]]
return sum
# Slight differences are possible due to the float/double precision.
self.assertAlmostEqual(score(root1.mapping), score(root2.mapping),
places=3)
def test_ordering_on(self, data):
def p(val, obj=None):
self.assert_(val >= 0 and val <= 100)
self.assertIsInstance(val, float)
matrix = instance_distance_matrix(data,
Euclidean(),
progress_callback=p)
root1 = HierarchicalClustering(matrix, progress_callback=p)
root2 = hier.clone(root1)
order_leaves_py(root1, matrix, progress_callback=p)
order_leaves_cpp(root2, matrix, progress_callback=p)
def score(mapping):
sum = 0.0
for i in range(matrix.dim - 1):
sum += matrix[mapping[i], mapping[i + 1]]
return sum
# Slight differences are possible due to the float/double precision.
self.assertAlmostEqual(score(root1.mapping),
score(root2.mapping),
places=3)
def test_clone(self):
cloned_cluster = hier.clone(self.cluster)
self.assertTrue(self.cluster.mapping.objects is cloned_cluster.mapping.objects)
self.assertEqual(self.cluster.mapping, cloned_cluster.mapping)
self.assertEqual(list(self.cluster), list(cloned_cluster))
def test_clone(self):
cloned_cluster = hier.clone(self.cluster)
self.assertTrue(
self.cluster.mapping.objects is cloned_cluster.mapping.objects)
self.assertEqual(self.cluster.mapping, cloned_cluster.mapping)
self.assertEqual(list(self.cluster), list(cloned_cluster))
