|
1
|
|
|
'''
|
|
2
|
|
|
|
|
3
|
|
|
Unit-tests for KD-Tree.
|
|
4
|
|
|
|
|
5
|
|
|
Copyright (C) 2015 Andrei Novikov ([email protected])
|
|
6
|
|
|
|
|
7
|
|
|
pyclustering is free software: you can redistribute it and/or modify
|
|
8
|
|
|
it under the terms of the GNU General Public License as published by
|
|
9
|
|
|
the Free Software Foundation, either version 3 of the License, or
|
|
10
|
|
|
(at your option) any later version.
|
|
11
|
|
|
|
|
12
|
|
|
pyclustering is distributed in the hope that it will be useful,
|
|
13
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
14
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
15
|
|
|
GNU General Public License for more details.
|
|
16
|
|
|
|
|
17
|
|
|
You should have received a copy of the GNU General Public License
|
|
18
|
|
|
along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
19
|
|
|
|
|
20
|
|
|
'''
|
|
21
|
|
|
|
|
22
|
|
|
import unittest;
|
|
23
|
|
|
import itertools;
|
|
24
|
|
|
|
|
25
|
|
|
from pyclustering.container.kdtree import kdtree;
|
|
26
|
|
|
|
|
27
|
|
|
class Test(unittest.TestCase):
|
|
28
|
|
|
def testKDTreeCreateWithoutPayload(self):
|
|
29
|
|
|
# Create k-d tree without any payload
|
|
30
|
|
|
array = [ [4, 3], [3, 4], [5, 8], [3, 3], [3, 9], [6, 4], [5, 9] ];
|
|
31
|
|
|
tree = kdtree(array);
|
|
32
|
|
|
|
|
33
|
|
|
assert len(tree.traverse()) == len(array);
|
|
34
|
|
|
for item in array:
|
|
35
|
|
|
node = tree.find_node(item);
|
|
36
|
|
|
|
|
37
|
|
|
assert node != None; # node should exist in the tree.
|
|
38
|
|
|
assert node.payload == None; # because we have created tree without any payloads.
|
|
39
|
|
|
assert node.data == item; # check for valid data.
|
|
40
|
|
|
|
|
41
|
|
|
|
|
42
|
|
|
def testKDTreeCreateWithPayload(self):
|
|
43
|
|
|
# Create k-d tree with payload
|
|
44
|
|
|
array = [ [4, 3], [3, 4], [5, 8], [3, 3], [3, 9], [6, 4], [5, 9] ];
|
|
45
|
|
|
payload = ['q', 'w', 'e', 'r', 't', 'y', 'u'];
|
|
46
|
|
|
|
|
47
|
|
|
tree = kdtree(array, payload);
|
|
48
|
|
|
assert len(tree.traverse()) == len(array);
|
|
49
|
|
|
for index in range(len(array)):
|
|
50
|
|
|
node = tree.find_node(array[index]);
|
|
51
|
|
|
|
|
52
|
|
|
assert node != None;
|
|
53
|
|
|
assert node.payload == payload[index];
|
|
54
|
|
|
assert node.data == array[index];
|
|
55
|
|
|
|
|
56
|
|
|
|
|
57
|
|
|
def testKDTreeCreateTrivial(self):
|
|
58
|
|
|
"Create k-d tree"
|
|
59
|
|
|
array = [ [3, 4], [5, 6], [9, 8], [7, 3], [1, 2], [2, 4], [2, 5], [3, 2] ];
|
|
60
|
|
|
tree = kdtree(array);
|
|
61
|
|
|
|
|
62
|
|
|
assert len(tree.traverse()) == len(array);
|
|
63
|
|
|
for item in array:
|
|
64
|
|
|
assert tree.find_node(item).data == item;
|
|
65
|
|
|
|
|
66
|
|
|
|
|
67
|
|
|
def testKDTreeInsertNodes(self):
|
|
68
|
|
|
"Create empty k-d tree and insert nodes"
|
|
69
|
|
|
array = [ [4, 3], [3, 4], [5, 8], [3, 3], [3, 9], [6, 4], [5, 9] ];
|
|
70
|
|
|
payload = ['q', 'w', 'e', 'r', 't', 'y', 'u'];
|
|
71
|
|
|
|
|
72
|
|
|
tree = kdtree();
|
|
73
|
|
|
assert len(tree.traverse()) == 0;
|
|
74
|
|
|
for index in range(len(array)):
|
|
75
|
|
|
node = tree.insert(array[index], payload[index]);
|
|
76
|
|
|
|
|
77
|
|
|
assert len(tree.traverse()) == index + 1;
|
|
78
|
|
|
|
|
79
|
|
|
assert node != None;
|
|
80
|
|
|
assert node.payload == payload[index];
|
|
81
|
|
|
assert node.data == array[index];
|
|
82
|
|
|
|
|
83
|
|
|
|
|
84
|
|
|
def testKDTreeParentSearch(self):
|
|
85
|
|
|
"Check for right parents"
|
|
86
|
|
|
array = [ [4, 3], [3, 4], [5, 8], [3, 3], [3, 9], [6, 4], [5, 9] ];
|
|
87
|
|
|
tree = kdtree(array);
|
|
88
|
|
|
|
|
89
|
|
|
node = tree.find_node([4, 3]);
|
|
90
|
|
|
assert node.parent == None;
|
|
91
|
|
|
|
|
92
|
|
|
node = tree.find_node([3, 4]);
|
|
93
|
|
|
assert node.parent.data == [4, 3];
|
|
94
|
|
|
|
|
95
|
|
|
node = tree.find_node([5, 8]);
|
|
96
|
|
|
assert node.parent.data == [4, 3];
|
|
97
|
|
|
|
|
98
|
|
|
node = tree.find_node([6, 4]);
|
|
99
|
|
|
assert node.parent.data == [5, 8];
|
|
100
|
|
|
|
|
101
|
|
|
node = tree.find_node([3, 3]);
|
|
102
|
|
|
assert node.parent.data == [3, 4];
|
|
103
|
|
|
|
|
104
|
|
|
node = tree.find_node([5, 9]);
|
|
105
|
|
|
assert node.parent.data == [5, 8];
|
|
106
|
|
|
|
|
107
|
|
|
node = tree.find_node([3, 9]);
|
|
108
|
|
|
assert node.parent.data == [3, 4];
|
|
109
|
|
|
|
|
110
|
|
|
|
|
111
|
|
|
def testKDTreeInsertRemoveNode1(self):
|
|
112
|
|
|
"Create empty k-d tree and insert nodes and after that remove all nodes"
|
|
113
|
|
|
array = [ [4, 3], [3, 4], [5, 8], [3, 3], [3, 9], [6, 4], [5, 9] ];
|
|
114
|
|
|
payload = ['q', 'w', 'e', 'r', 't', 'y', 'u'];
|
|
115
|
|
|
|
|
116
|
|
|
tree = kdtree();
|
|
117
|
|
|
for index in range(len(array)):
|
|
118
|
|
|
node = tree.insert(array[index], payload[index]);
|
|
119
|
|
|
|
|
120
|
|
|
length = len(array);
|
|
121
|
|
|
for index in range(0, length):
|
|
122
|
|
|
node = tree.remove(array[index]);
|
|
123
|
|
|
assert len(tree.traverse()) == length - index - 1;
|
|
124
|
|
|
|
|
125
|
|
|
if (index + 1 < length): # When root is removed then None will be returned
|
|
126
|
|
|
assert node != None;
|
|
127
|
|
|
else:
|
|
128
|
|
|
assert node == None;
|
|
129
|
|
|
|
|
130
|
|
|
# Check other nodes are located in the tree
|
|
131
|
|
|
for k in range(index + 1, length):
|
|
132
|
|
|
node = tree.find_node(array[k]);
|
|
133
|
|
|
|
|
134
|
|
|
assert node.data == array[k];
|
|
135
|
|
|
assert node.payload == payload[k];
|
|
136
|
|
|
|
|
137
|
|
|
|
|
138
|
|
|
def testKDTreeInsertRemoveNode2(self):
|
|
139
|
|
|
# This test simulates situation when a bug (16.01.2014) with removing was occuring
|
|
140
|
|
|
array = [ [9, 9], [3, 3], [4, 4] ];
|
|
141
|
|
|
tree = kdtree(array);
|
|
142
|
|
|
|
|
143
|
|
|
assert None != tree.remove([9, 9]);
|
|
144
|
|
|
assert len(tree.traverse()) == 2;
|
|
145
|
|
|
|
|
146
|
|
|
assert None != tree.remove([4, 4]);
|
|
147
|
|
|
assert len(tree.traverse()) == 1;
|
|
148
|
|
|
|
|
149
|
|
|
assert None == tree.remove([3, 3]);
|
|
150
|
|
|
assert len(tree.traverse()) == 0;
|
|
151
|
|
|
|
|
152
|
|
|
|
|
153
|
|
|
def testKDTreeRemoveLongBranch(self):
|
|
154
|
|
|
# Create only one branch - worth case and remove it
|
|
155
|
|
|
array = [ [5, 5], [6, 5], [6, 6], [7, 6], [7, 7] ];
|
|
156
|
|
|
tree = kdtree(array);
|
|
157
|
|
|
|
|
158
|
|
|
assert len(tree.traverse()) == len(array);
|
|
159
|
|
|
#tree.show();
|
|
160
|
|
|
|
|
161
|
|
|
for index in range(len(array)):
|
|
162
|
|
|
node = tree.remove(array[index]);
|
|
163
|
|
|
assert len(tree.traverse()) == len(array) - index - 1;
|
|
164
|
|
|
|
|
165
|
|
|
# Remove from other end
|
|
166
|
|
|
tree = kdtree(array);
|
|
167
|
|
|
for index in range(len(array)):
|
|
168
|
|
|
node = tree.remove(array[len(array) - index - 1]);
|
|
169
|
|
|
assert len(tree.traverse()) == len(array) - index - 1;
|
|
170
|
|
|
|
|
171
|
|
|
|
|
172
|
|
|
def testKDTreeNearestNodeTrivial1(self):
|
|
173
|
|
|
array = [ [4, 3], [3, 4], [5, 8], [3, 3], [3, 9], [6, 4], [6, 9], [4, 9] ];
|
|
174
|
|
|
tree = kdtree(array);
|
|
175
|
|
|
|
|
176
|
|
|
for item in array:
|
|
177
|
|
|
assert tree.find_nearest_dist_node(item, 0).data == item;
|
|
178
|
|
|
assert tree.find_nearest_dist_node(item, 0.5).data == item;
|
|
179
|
|
|
assert tree.find_nearest_dist_node(item, 1).data == item;
|
|
180
|
|
|
assert tree.find_nearest_dist_node(item, 3).data == item;
|
|
181
|
|
|
assert tree.find_nearest_dist_node(item, 10).data == item;
|
|
182
|
|
|
|
|
183
|
|
|
assert tree.find_nearest_dist_node([6.1, 4.1], 0.5).data == [6, 4];
|
|
184
|
|
|
assert tree.find_nearest_dist_node([6, 12], 0) == None;
|
|
185
|
|
|
assert tree.find_nearest_dist_node([6, 12], 1) == None;
|
|
186
|
|
|
assert tree.find_nearest_dist_node([6, 12], 3).data == [6, 9];
|
|
187
|
|
|
|
|
188
|
|
|
|
|
189
|
|
|
def testKDTreeNearestNodeTrivial2(self):
|
|
190
|
|
|
arrays = [
|
|
191
|
|
|
[ [3, 4], [5, 6], [9, 8], [7, 3], [1, 2], [2, 4], [2, 5], [3, 2], [3, 3] ],
|
|
192
|
|
|
[ [5, 6], [1, 3], [7, 3], [1, 1], [9, 9], [4, 7], [0, 3], [3, 5], [1, 2], [9, 3], [9, 8], [5, 5], [6, 6], [0, 0], [-4, -5], [-1, 5], [-8, 3] ]
|
|
193
|
|
|
];
|
|
194
|
|
|
|
|
195
|
|
|
distances = [0.0, 0.5, 1.0, 3.0, 10.0];
|
|
196
|
|
|
|
|
197
|
|
|
for array in arrays:
|
|
198
|
|
|
tree = kdtree(array);
|
|
199
|
|
|
|
|
200
|
|
|
for item in array:
|
|
201
|
|
|
for distance in distances:
|
|
202
|
|
|
assert tree.find_nearest_dist_node(item, distance).data == item;
|
|
203
|
|
|
|
|
204
|
|
|
# Verification test, so it is required too much time for testing.
|
|
205
|
|
|
#def testVerificationKdTree1(self):
|
|
206
|
|
|
# array = [ [5, 5], [4, 5], [4, 4], [3, 4], [3, 3], [6, 6], [8, 8], [7, 7], [9, 9]];
|
|
207
|
|
|
# self.template_verification_insert_remove_kdtree_test(array);
|
|
208
|
|
|
|
|
209
|
|
|
def templateVerificationInsertRemoveKDTreeTest(self, array):
|
|
210
|
|
|
for perm_array in itertools.permutations(array):
|
|
211
|
|
|
tree = kdtree(array);
|
|
212
|
|
|
length = len(array);
|
|
213
|
|
|
|
|
214
|
|
|
for index in range(len(perm_array)):
|
|
215
|
|
|
#tree.show();
|
|
216
|
|
|
|
|
217
|
|
|
node = tree.remove(perm_array[index]);
|
|
218
|
|
|
|
|
219
|
|
|
if ( index + 1 < length ):
|
|
220
|
|
|
assert node is not None;
|
|
221
|
|
|
|
|
222
|
|
|
assert len(tree.traverse()) == length - index - 1;
|
|
223
|
|
|
|
|
224
|
|
|
|
|
225
|
|
|
if __name__ == "__main__":
|
|
226
|
|
|
unittest.main(); |
annoviko /
pyclustering
