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
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
|
package de.lmu.ifi.dbs.elki.data.synthetic.bymodel;
/*
This file is part of ELKI:
Environment for Developing KDD-Applications Supported by Index-Structures
Copyright (C) 2011
Ludwig-Maximilians-Universität München
Lehr- und Forschungseinheit für Datenbanksysteme
ELKI Development Team
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
import java.util.ArrayList;
import java.util.Collections;
import java.util.LinkedList;
import java.util.List;
import de.lmu.ifi.dbs.elki.data.ClassLabel;
import de.lmu.ifi.dbs.elki.data.DoubleVector;
import de.lmu.ifi.dbs.elki.data.SimpleClassLabel;
import de.lmu.ifi.dbs.elki.data.type.TypeUtil;
import de.lmu.ifi.dbs.elki.data.type.VectorFieldTypeInformation;
import de.lmu.ifi.dbs.elki.datasource.bundle.MultipleObjectsBundle;
import de.lmu.ifi.dbs.elki.math.linearalgebra.Vector;
import de.lmu.ifi.dbs.elki.utilities.exceptions.UnableToComplyException;
/**
* Generate a data set according to a given model.
*
* Key idea of this generator is to re-generate points if they are more likely
* to belong to a different cluster than the one they were generated for. The
* benefit is that we should end up with a data set that follows closely the
* model that we specified.
*
* The drawbacks are that on one hand, specifications might be unsatisfiable.
* For this a retry count is kept and an {@link UnableToComplyException} is
* thrown when the maximum number of retries is exceeded.
*
* On the other hand, the model might not be exactly as specified. When the
* generator reports an "Density correction factor estimation" that differs from
* 1.0 this is an indication that the result is not exact.
*
* On the third hand, rejecting points introduces effects where one generator
* can influence others, so random generator results will not be stable with
* respect to the addition of new dimensions and similar if there are any
* rejects involved. So this generator is not entirely optimal for generating
* data sets for scalability tests on the number of dimensions, although if
* clusters overlap little enough (so that no rejects happen) the results should
* be as expected.
*
* @author Erich Schubert
*
* @apiviz.has GeneratorInterface
*/
public class GeneratorMain {
/**
* List of clusters to generate
*/
private LinkedList<GeneratorInterface> generators = new LinkedList<GeneratorInterface>();
/**
* Add a cluster to the cluster list.
*
* @param c cluster to add
*/
public void addCluster(GeneratorInterface c) {
generators.add(c);
}
/**
* Controls whether points are tested against the model during generation
*/
private boolean testAgainstModel = true;
/**
* Main loop to generate data set.
*
* @throws UnableToComplyException when model not satisfiable or no clusters
* specified.
*/
public void generate() throws UnableToComplyException {
// we actually need some clusters.
if(generators.size() < 1) {
throw new UnableToComplyException("No clusters specified.");
}
// Assert that cluster dimensions agree.
final int dim = generators.get(0).getDim();
for(GeneratorInterface c : generators) {
if(c.getDim() != dim) {
throw new UnableToComplyException("Cluster dimensions do not agree.");
}
}
// generate clusters
for(GeneratorInterface curclus : generators) {
while(curclus.getPoints().size() < curclus.getSize()) {
// generate the "missing" number of points
List<Vector> newp = curclus.generate(curclus.getSize() - curclus.getPoints().size());
if(curclus instanceof GeneratorInterfaceDynamic) {
GeneratorInterfaceDynamic cursclus = (GeneratorInterfaceDynamic) curclus;
for(Vector p : newp) {
if(testAgainstModel) {
double max = 0.0;
double is = 0.0;
for(GeneratorInterface other : generators) {
double d = other.getDensity(p) * other.getSize();
if(other == curclus) {
is = d;
}
else if(d > max) {
max = d;
}
}
// Only keep the point if the largest density was the cluster it
// was generated for
if(is >= max) {
cursclus.getPoints().add(p);
}
else {
cursclus.addDiscarded(1);
}
}
else {
cursclus.getPoints().add(p);
}
}
}
}
}
}
/**
* Return value of the {@link #testAgainstModel} flag
*
* @return value of testAgainstModel
*/
public boolean isTestAgainstModel() {
return testAgainstModel;
}
/**
* Set the value of the {@link #testAgainstModel} flag
*
* @param testAgainstModel New value
*/
public void setTestAgainstModel(boolean testAgainstModel) {
this.testAgainstModel = testAgainstModel;
}
/**
* Access the generators.
*
* @return generators
*/
public List<GeneratorInterface> getGenerators() {
return Collections.unmodifiableList(generators);
}
/**
* Get the objects bundle
*
* @return Bundle
*/
public MultipleObjectsBundle getBundle() {
final int dim = generators.get(0).getDim();
final DoubleVector factory = new DoubleVector(new double[dim]);
MultipleObjectsBundle bundle = new MultipleObjectsBundle();
VectorFieldTypeInformation<DoubleVector> type = new VectorFieldTypeInformation<DoubleVector>(DoubleVector.class, dim, factory);
bundle.appendColumn(type, new ArrayList<Object>());
bundle.appendColumn(TypeUtil.CLASSLABEL, new ArrayList<Object>());
for(GeneratorInterface generator : generators) {
ClassLabel l = new SimpleClassLabel(generator.getName());
for(Vector v : generator.getPoints()) {
DoubleVector dv = new DoubleVector(v);
bundle.appendSimple(dv, l);
}
}
return bundle;
}
}
|
