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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
|
package de.lmu.ifi.dbs.elki.index.preprocessed.localpca;
/*
This file is part of ELKI:
Environment for Developing KDD-Applications Supported by Index-Structures
Copyright (C) 2012
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 de.lmu.ifi.dbs.elki.data.NumberVector;
import de.lmu.ifi.dbs.elki.data.type.TypeInformation;
import de.lmu.ifi.dbs.elki.database.datastore.DataStoreFactory;
import de.lmu.ifi.dbs.elki.database.datastore.DataStoreUtil;
import de.lmu.ifi.dbs.elki.database.ids.DBIDIter;
import de.lmu.ifi.dbs.elki.database.ids.DBIDRef;
import de.lmu.ifi.dbs.elki.database.relation.Relation;
import de.lmu.ifi.dbs.elki.distance.distancefunction.DistanceFunction;
import de.lmu.ifi.dbs.elki.distance.distancefunction.EuclideanDistanceFunction;
import de.lmu.ifi.dbs.elki.distance.distanceresultlist.DistanceDBIDResult;
import de.lmu.ifi.dbs.elki.distance.distancevalue.DoubleDistance;
import de.lmu.ifi.dbs.elki.index.preprocessed.AbstractPreprocessorIndex;
import de.lmu.ifi.dbs.elki.logging.progress.FiniteProgress;
import de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCAFilteredResult;
import de.lmu.ifi.dbs.elki.math.linearalgebra.pca.PCAFilteredRunner;
import de.lmu.ifi.dbs.elki.utilities.ClassGenericsUtil;
import de.lmu.ifi.dbs.elki.utilities.documentation.Description;
import de.lmu.ifi.dbs.elki.utilities.documentation.Title;
import de.lmu.ifi.dbs.elki.utilities.exceptions.ExceptionMessages;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.AbstractParameterizer;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.OptionID;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.Parameterizable;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.parameterization.Parameterization;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.parameters.ObjectParameter;
/**
* Abstract base class for a local PCA based index.
*
* @author Elke Achtert
* @author Erich Schubert
*
* @apiviz.has PCAFilteredRunner
*
* @param <NV> Vector type
*/
// TODO: loosen DoubleDistance restriction.
@Title("Local PCA Preprocessor")
@Description("Materializes the local PCA and the locally weighted matrix of objects of a database.")
public abstract class AbstractFilteredPCAIndex<NV extends NumberVector<?>> extends AbstractPreprocessorIndex<NV, PCAFilteredResult> implements FilteredLocalPCAIndex<NV> {
/**
* PCA utility object.
*/
protected final PCAFilteredRunner<NV> pca;
/**
* Constructor.
*
* @param relation Relation to use
* @param pca PCA runner to use
*/
public AbstractFilteredPCAIndex(Relation<NV> relation, PCAFilteredRunner<NV> pca) {
super(relation);
this.pca = pca;
}
/**
* Preprocessing step.
*/
protected void preprocess() {
if(relation == null || relation.size() <= 0) {
throw new IllegalArgumentException(ExceptionMessages.DATABASE_EMPTY);
}
// Note: this is required for ERiC to work properly, otherwise the data is
// recomputed for the partitions!
if(storage != null) {
return;
}
storage = DataStoreUtil.makeStorage(relation.getDBIDs(), DataStoreFactory.HINT_HOT | DataStoreFactory.HINT_TEMP, PCAFilteredResult.class);
long start = System.currentTimeMillis();
FiniteProgress progress = getLogger().isVerbose() ? new FiniteProgress("Performing local PCA", relation.size(), getLogger()) : null;
// TODO: use a bulk operation?
for(DBIDIter iditer = relation.iterDBIDs(); iditer.valid(); iditer.advance()) {
DistanceDBIDResult<DoubleDistance> objects = objectsForPCA(iditer);
PCAFilteredResult pcares = pca.processQueryResult(objects, relation);
storage.put(iditer, pcares);
if(progress != null) {
progress.incrementProcessed(getLogger());
}
}
if(progress != null) {
progress.ensureCompleted(getLogger());
}
long end = System.currentTimeMillis();
if(getLogger().isVerbose()) {
long elapsedTime = end - start;
getLogger().verbose(this.getClass().getName() + " runtime: " + elapsedTime + " milliseconds.");
}
}
@Override
public PCAFilteredResult getLocalProjection(DBIDRef objid) {
if(storage == null) {
preprocess();
}
return storage.get(objid);
}
/**
* Returns the objects to be considered within the PCA for the specified query
* object.
*
* @param id the id of the query object for which a PCA should be performed
* @return the list of the objects (i.e. the ids and the distances to the
* query object) to be considered within the PCA
*/
protected abstract DistanceDBIDResult<DoubleDistance> objectsForPCA(DBIDRef id);
/**
* Factory class.
*
* @author Erich Schubert
*
* @apiviz.stereotype factory
* @apiviz.uses AbstractFilteredPCAIndex oneway - - «create»
*/
public abstract static class Factory<NV extends NumberVector<?>, I extends AbstractFilteredPCAIndex<NV>> implements FilteredLocalPCAIndex.Factory<NV, I>, Parameterizable {
/**
* Parameter to specify the distance function used for running PCA.
*
* Key: {@code -localpca.distancefunction}
*/
public static final OptionID PCA_DISTANCE_ID = new OptionID("localpca.distancefunction", "The distance function used to select objects for running PCA.");
/**
* Holds the instance of the distance function specified by
* {@link #PCA_DISTANCE_ID}.
*/
protected DistanceFunction<NV, DoubleDistance> pcaDistanceFunction;
/**
* PCA utility object.
*/
protected PCAFilteredRunner<NV> pca;
/**
* Constructor.
*
* @param pcaDistanceFunction distance Function
* @param pca PCA runner
*/
public Factory(DistanceFunction<NV, DoubleDistance> pcaDistanceFunction, PCAFilteredRunner<NV> pca) {
super();
this.pcaDistanceFunction = pcaDistanceFunction;
this.pca = pca;
}
@Override
public abstract I instantiate(Relation<NV> relation);
@Override
public TypeInformation getInputTypeRestriction() {
return pcaDistanceFunction.getInputTypeRestriction();
}
/**
* Parameterization class.
*
* @author Erich Schubert
*
* @apiviz.exclude
*/
public abstract static class Parameterizer<NV extends NumberVector<?>, I extends AbstractFilteredPCAIndex<NV>> extends AbstractParameterizer {
/**
* Holds the instance of the distance function specified by
* {@link #PCA_DISTANCE_ID}.
*/
protected DistanceFunction<NV, DoubleDistance> pcaDistanceFunction;
/**
* PCA utility object.
*/
protected PCAFilteredRunner<NV> pca;
@Override
protected void makeOptions(Parameterization config) {
super.makeOptions(config);
final ObjectParameter<DistanceFunction<NV, DoubleDistance>> pcaDistanceFunctionP = new ObjectParameter<DistanceFunction<NV, DoubleDistance>>(PCA_DISTANCE_ID, DistanceFunction.class, EuclideanDistanceFunction.class);
if(config.grab(pcaDistanceFunctionP)) {
pcaDistanceFunction = pcaDistanceFunctionP.instantiateClass(config);
}
Class<PCAFilteredRunner<NV>> cls = ClassGenericsUtil.uglyCastIntoSubclass(PCAFilteredRunner.class);
pca = config.tryInstantiate(cls);
}
}
}
}
|