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
|
package de.lmu.ifi.dbs.elki.index.tree.spatial.rstarvariants.query;
/*
This file is part of ELKI:
Environment for Developing KDD-Applications Supported by Index-Structures
Copyright (C) 2015
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.HashMap;
import java.util.List;
import java.util.Map;
import de.lmu.ifi.dbs.elki.data.NumberVector;
import de.lmu.ifi.dbs.elki.database.QueryUtil;
import de.lmu.ifi.dbs.elki.database.ids.ArrayDBIDs;
import de.lmu.ifi.dbs.elki.database.ids.DBID;
import de.lmu.ifi.dbs.elki.database.ids.DBIDIter;
import de.lmu.ifi.dbs.elki.database.ids.DBIDUtil;
import de.lmu.ifi.dbs.elki.database.ids.KNNHeap;
import de.lmu.ifi.dbs.elki.database.ids.KNNList;
import de.lmu.ifi.dbs.elki.database.relation.Relation;
import de.lmu.ifi.dbs.elki.distance.distancefunction.minkowski.EuclideanDistanceFunction;
import de.lmu.ifi.dbs.elki.distance.distancefunction.minkowski.SquaredEuclideanDistanceFunction;
import de.lmu.ifi.dbs.elki.index.tree.query.DoubleDistanceSearchCandidate;
import de.lmu.ifi.dbs.elki.index.tree.spatial.SpatialDirectoryEntry;
import de.lmu.ifi.dbs.elki.index.tree.spatial.SpatialPointLeafEntry;
import de.lmu.ifi.dbs.elki.index.tree.spatial.rstarvariants.AbstractRStarTree;
import de.lmu.ifi.dbs.elki.index.tree.spatial.rstarvariants.AbstractRStarTreeNode;
import de.lmu.ifi.dbs.elki.utilities.datastructures.heap.ComparableMinHeap;
import de.lmu.ifi.dbs.elki.utilities.documentation.Reference;
/**
* Instance of a KNN query for a particular spatial index.
*
* Reference:
* <p>
* G. R. Hjaltason, H. Samet<br />
* Ranking in spatial databases<br />
* In: 4th Symposium on Advances in Spatial Databases, SSD'95
* </p>
*
* @author Erich Schubert
* @since 0.7.0
*
* @apiviz.uses EuclideanDistanceFunction
* @apiviz.uses SquaredEuclideanDistanceFunction
*/
@Reference(authors = "G. R. Hjaltason, H. Samet", //
title = "Ranking in spatial databases", //
booktitle = "Advances in Spatial Databases - 4th Symposium, SSD'95", //
url = "http://dx.doi.org/10.1007/3-540-60159-7_6")
public class EuclideanRStarTreeKNNQuery<O extends NumberVector> extends RStarTreeKNNQuery<O> {
/**
* Squared euclidean distance function.
*/
private static final SquaredEuclideanDistanceFunction SQUARED = SquaredEuclideanDistanceFunction.STATIC;
/**
* Constructor.
*
* @param tree Index to use
* @param relation Data relation to query
*/
public EuclideanRStarTreeKNNQuery(AbstractRStarTree<?, ?, ?> tree, Relation<? extends O> relation) {
super(tree, relation, EuclideanDistanceFunction.STATIC);
}
@Override
public KNNList getKNNForObject(O obj, int k) {
if(k < 1) {
throw new IllegalArgumentException("At least one neighbor has to be requested!");
}
tree.statistics.countKNNQuery();
final KNNHeap knnList = DBIDUtil.newHeap(k);
final ComparableMinHeap<DoubleDistanceSearchCandidate> pq = new ComparableMinHeap<>(Math.min(knnList.getK() << 1, 21));
// expand root
double maxDist = expandNode(obj, knnList, pq, Double.MAX_VALUE, tree.getRootID());
// search in tree
while(!pq.isEmpty()) {
DoubleDistanceSearchCandidate pqNode = pq.poll();
if(pqNode.mindist > maxDist) {
break;
}
maxDist = expandNode(obj, knnList, pq, maxDist, pqNode.nodeID);
}
return QueryUtil.applySqrt(knnList.toKNNList());
}
private double expandNode(O object, KNNHeap knnList, final ComparableMinHeap<DoubleDistanceSearchCandidate> pq, double maxDist, final int nodeID) {
AbstractRStarTreeNode<?, ?> node = tree.getNode(nodeID);
// data node
if(node.isLeaf()) {
for(int i = 0; i < node.getNumEntries(); i++) {
SpatialPointLeafEntry entry = (SpatialPointLeafEntry) node.getEntry(i);
double distance = SQUARED.minDist(entry, object);
tree.statistics.countDistanceCalculation();
if(distance <= maxDist) {
maxDist = knnList.insert(distance, entry.getDBID());
}
}
}
// directory node
else {
for(int i = 0; i < node.getNumEntries(); i++) {
SpatialDirectoryEntry entry = (SpatialDirectoryEntry) node.getEntry(i);
double distance = SQUARED.minDist(entry, object);
tree.statistics.countDistanceCalculation();
// Greedy expand, bypassing the queue
if(distance <= 0) {
expandNode(object, knnList, pq, maxDist, entry.getPageID());
}
else {
if(distance <= maxDist) {
pq.add(new DoubleDistanceSearchCandidate(distance, entry.getPageID()));
}
}
}
}
return maxDist;
}
@Override
public List<KNNList> getKNNForBulkDBIDs(ArrayDBIDs ids, int k) {
if(k < 1) {
throw new IllegalArgumentException("At least one enumeration has to be requested!");
}
// While this works, it seems to be slow at least for large sets!
// TODO: use a DataStore instead of a map.
final Map<DBID, KNNHeap> knnLists = new HashMap<>(ids.size());
for(DBIDIter iter = ids.iter(); iter.valid(); iter.advance()) {
DBID id = DBIDUtil.deref(iter);
knnLists.put(id, DBIDUtil.newHeap(k));
}
batchNN(tree.getRoot(), knnLists);
List<KNNList> result = new ArrayList<>();
for(DBIDIter iter = ids.iter(); iter.valid(); iter.advance()) {
DBID id = DBIDUtil.deref(iter);
tree.statistics.countKNNQuery();
result.add(QueryUtil.applySqrt(knnLists.get(id).toKNNList()));
}
return result;
}
}
|