1 | /* |
---|
2 | * This program is free software; you can redistribute it and/or modify |
---|
3 | * it under the terms of the GNU General Public License as published by |
---|
4 | * the Free Software Foundation; either version 2 of the License, or |
---|
5 | * (at your option) any later version. |
---|
6 | * |
---|
7 | * This program is distributed in the hope that it will be useful, |
---|
8 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
9 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
---|
10 | * GNU General Public License for more details. |
---|
11 | * |
---|
12 | * You should have received a copy of the GNU General Public License |
---|
13 | * along with this program; if not, write to the Free Software |
---|
14 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. |
---|
15 | */ |
---|
16 | |
---|
17 | /* |
---|
18 | * MidPointOfWidestDimension.java |
---|
19 | * Copyright (C) 2007 University of Waikato, Hamilton, New Zealand |
---|
20 | */ |
---|
21 | |
---|
22 | package weka.core.neighboursearch.kdtrees; |
---|
23 | |
---|
24 | import weka.core.RevisionUtils; |
---|
25 | import weka.core.TechnicalInformation; |
---|
26 | import weka.core.TechnicalInformationHandler; |
---|
27 | import weka.core.TechnicalInformation.Field; |
---|
28 | import weka.core.TechnicalInformation.Type; |
---|
29 | |
---|
30 | /** |
---|
31 | <!-- globalinfo-start --> |
---|
32 | * The class that splits a KDTree node based on the midpoint value of a dimension in which the node's points have the widest spread.<br/> |
---|
33 | * <br/> |
---|
34 | * For more information see also:<br/> |
---|
35 | * <br/> |
---|
36 | * Andrew Moore (1991). A tutorial on kd-trees. |
---|
37 | * <p/> |
---|
38 | <!-- globalinfo-end --> |
---|
39 | * |
---|
40 | <!-- technical-bibtex-start --> |
---|
41 | * BibTeX: |
---|
42 | * <pre> |
---|
43 | * @techreport{Moore1991, |
---|
44 | * author = {Andrew Moore}, |
---|
45 | * booktitle = {University of Cambridge Computer Laboratory Technical Report No. 209}, |
---|
46 | * howpublished = {Extract from PhD Thesis}, |
---|
47 | * title = {A tutorial on kd-trees}, |
---|
48 | * year = {1991}, |
---|
49 | * HTTP = {http://www.autonlab.org/autonweb/14665.html} |
---|
50 | * } |
---|
51 | * </pre> |
---|
52 | * <p/> |
---|
53 | <!-- technical-bibtex-end --> |
---|
54 | * |
---|
55 | <!-- options-start --> |
---|
56 | <!-- options-end --> |
---|
57 | * |
---|
58 | * @author Ashraf M. Kibriya (amk14[at-the-rate]cs[dot]waikato[dot]ac[dot]nz) |
---|
59 | * @version $Revision: 5953 $ |
---|
60 | */ |
---|
61 | public class MidPointOfWidestDimension |
---|
62 | extends KDTreeNodeSplitter |
---|
63 | implements TechnicalInformationHandler { |
---|
64 | |
---|
65 | /** for serialization. */ |
---|
66 | private static final long serialVersionUID = -7617277960046591906L; |
---|
67 | |
---|
68 | /** |
---|
69 | * Returns a string describing this nearest neighbour search algorithm. |
---|
70 | * |
---|
71 | * @return a description of the algorithm for displaying in the |
---|
72 | * explorer/experimenter gui |
---|
73 | */ |
---|
74 | public String globalInfo() { |
---|
75 | return |
---|
76 | "The class that splits a KDTree node based on the midpoint value of " |
---|
77 | + "a dimension in which the node's points have the widest spread.\n\n" |
---|
78 | + "For more information see also:\n\n" |
---|
79 | + getTechnicalInformation().toString(); |
---|
80 | } |
---|
81 | |
---|
82 | /** |
---|
83 | * Returns an instance of a TechnicalInformation object, containing detailed |
---|
84 | * information about the technical background of this class, e.g., paper |
---|
85 | * reference or book this class is based on. |
---|
86 | * |
---|
87 | * @return the technical information about this class |
---|
88 | */ |
---|
89 | public TechnicalInformation getTechnicalInformation() { |
---|
90 | TechnicalInformation result; |
---|
91 | |
---|
92 | result = new TechnicalInformation(Type.TECHREPORT); |
---|
93 | result.setValue(Field.AUTHOR, "Andrew Moore"); |
---|
94 | result.setValue(Field.YEAR, "1991"); |
---|
95 | result.setValue(Field.TITLE, "A tutorial on kd-trees"); |
---|
96 | result.setValue(Field.HOWPUBLISHED, "Extract from PhD Thesis"); |
---|
97 | result.setValue(Field.BOOKTITLE, "University of Cambridge Computer Laboratory Technical Report No. 209"); |
---|
98 | result.setValue(Field.HTTP, "http://www.autonlab.org/autonweb/14665.html"); |
---|
99 | |
---|
100 | return result; |
---|
101 | } |
---|
102 | |
---|
103 | /** |
---|
104 | * Splits a node into two based on the midpoint value of the dimension |
---|
105 | * in which the points have the widest spread. After splitting two |
---|
106 | * new nodes are created and correctly initialised. And, node.left |
---|
107 | * and node.right are set appropriately. |
---|
108 | * @param node The node to split. |
---|
109 | * @param numNodesCreated The number of nodes that so far have been |
---|
110 | * created for the tree, so that the newly created nodes are |
---|
111 | * assigned correct/meaningful node numbers/ids. |
---|
112 | * @param nodeRanges The attributes' range for the points inside |
---|
113 | * the node that is to be split. |
---|
114 | * @param universe The attributes' range for the whole |
---|
115 | * point-space. |
---|
116 | * @throws Exception If there is some problem in splitting the |
---|
117 | * given node. |
---|
118 | */ |
---|
119 | public void splitNode(KDTreeNode node, int numNodesCreated, |
---|
120 | double[][] nodeRanges, double[][] universe) throws Exception { |
---|
121 | |
---|
122 | correctlyInitialized(); |
---|
123 | |
---|
124 | int splitDim = widestDim(nodeRanges, universe); |
---|
125 | |
---|
126 | double splitVal = m_EuclideanDistance.getMiddle(nodeRanges[splitDim]); |
---|
127 | |
---|
128 | int rightStart = rearrangePoints(m_InstList, node.m_Start, node.m_End, |
---|
129 | splitDim, splitVal); |
---|
130 | |
---|
131 | if (rightStart == node.m_Start || rightStart > node.m_End) { |
---|
132 | if (rightStart == node.m_Start) |
---|
133 | throw new Exception("Left child is empty in node " |
---|
134 | + node.m_NodeNumber + |
---|
135 | ". Not possible with " + |
---|
136 | "MidPointofWidestDim splitting method. Please " + |
---|
137 | "check code."); |
---|
138 | else |
---|
139 | throw new Exception("Right child is empty in node " + node.m_NodeNumber + |
---|
140 | ". Not possible with " + |
---|
141 | "MidPointofWidestDim splitting method. Please " + |
---|
142 | "check code."); |
---|
143 | } |
---|
144 | |
---|
145 | node.m_SplitDim = splitDim; |
---|
146 | node.m_SplitValue = splitVal; |
---|
147 | node.m_Left = new KDTreeNode(numNodesCreated + 1, node.m_Start, |
---|
148 | rightStart - 1, m_EuclideanDistance.initializeRanges(m_InstList, |
---|
149 | node.m_Start, rightStart - 1)); |
---|
150 | node.m_Right = new KDTreeNode(numNodesCreated + 2, rightStart, node.m_End, |
---|
151 | m_EuclideanDistance |
---|
152 | .initializeRanges(m_InstList, rightStart, node.m_End)); |
---|
153 | } |
---|
154 | |
---|
155 | /** |
---|
156 | * Re-arranges the indices array such that the points <= to the splitVal |
---|
157 | * are on the left of the array and those > the splitVal are on the right. |
---|
158 | * |
---|
159 | * @param indices The master index array. |
---|
160 | * @param startidx The begining index of portion of indices that needs |
---|
161 | * re-arranging. |
---|
162 | * @param endidx The end index of portion of indices that needs |
---|
163 | * re-arranging. |
---|
164 | * @param splitDim The split dimension/attribute. |
---|
165 | * @param splitVal The split value. |
---|
166 | * @return The startIdx of the points > the splitVal (the points |
---|
167 | * belonging to the right child of the node). |
---|
168 | */ |
---|
169 | protected int rearrangePoints(int[] indices, final int startidx, final int endidx, |
---|
170 | final int splitDim, final double splitVal) { |
---|
171 | |
---|
172 | int tmp, left = startidx - 1; |
---|
173 | for (int i = startidx; i <= endidx; i++) { |
---|
174 | if (m_EuclideanDistance.valueIsSmallerEqual(m_Instances |
---|
175 | .instance(indices[i]), splitDim, splitVal)) { |
---|
176 | left++; |
---|
177 | tmp = indices[left]; |
---|
178 | indices[left] = indices[i]; |
---|
179 | indices[i] = tmp; |
---|
180 | }//end if valueIsSmallerEqual |
---|
181 | }//end for |
---|
182 | return left + 1; |
---|
183 | } |
---|
184 | |
---|
185 | /** |
---|
186 | * Returns the revision string. |
---|
187 | * |
---|
188 | * @return the revision |
---|
189 | */ |
---|
190 | public String getRevision() { |
---|
191 | return RevisionUtils.extract("$Revision: 5953 $"); |
---|
192 | } |
---|
193 | } |
---|