| 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 | * Test.java |
|---|
| 19 | * Copyright (C) 2000 University of Waikato, Hamilton, New Zealand |
|---|
| 20 | * |
|---|
| 21 | */ |
|---|
| 22 | |
|---|
| 23 | package weka.datagenerators; |
|---|
| 24 | |
|---|
| 25 | import weka.core.Attribute; |
|---|
| 26 | import weka.core.Instance; |
|---|
| 27 | import weka.core.Instances; |
|---|
| 28 | import weka.core.RevisionHandler; |
|---|
| 29 | import weka.core.RevisionUtils; |
|---|
| 30 | import weka.core.Utils; |
|---|
| 31 | |
|---|
| 32 | import java.io.Serializable; |
|---|
| 33 | |
|---|
| 34 | /** |
|---|
| 35 | * Class to represent a test. <br/> |
|---|
| 36 | * <br/> |
|---|
| 37 | * The string representation of the test can be supplied in standard notation |
|---|
| 38 | * or for a subset of types of attributes in Prolog notation.<br/> |
|---|
| 39 | * |
|---|
| 40 | * Following examples for all possible tests that can be represented by |
|---|
| 41 | * this class, given in standard notation.<br/> |
|---|
| 42 | * <br/> |
|---|
| 43 | * Examples of tests for numeric attributes:<br/> |
|---|
| 44 | * B >= 2.333<br/> B < 4.56<br/> |
|---|
| 45 | * <br/> |
|---|
| 46 | * Examples of tests for nominal attributes with more then 2 values:<br/> |
|---|
| 47 | * A = rain <br/> A != rain<br/> |
|---|
| 48 | * <br/> |
|---|
| 49 | * Examples of tests for nominal attribute with exactly 2 values:<br/> |
|---|
| 50 | * A = false <br/> A = true<br/> |
|---|
| 51 | * <br/> |
|---|
| 52 | * <br/> |
|---|
| 53 | * The Prolog notation is only supplied for numeric attributes and |
|---|
| 54 | * for nominal attributes that have the values "true" and "false".<br/> |
|---|
| 55 | * <br/> |
|---|
| 56 | * Following examples for the Prolog notation provided.<br/> |
|---|
| 57 | * <br/> |
|---|
| 58 | * Examples of tests for numeric attributes:<br/> |
|---|
| 59 | * The same as for standard notation above.<br/> |
|---|
| 60 | * <br/> |
|---|
| 61 | * Examples of tests for nominal attributes with values "true"and "false":<br/> |
|---|
| 62 | * A<br/> |
|---|
| 63 | * not(A)<br/> |
|---|
| 64 | * <br/> |
|---|
| 65 | * (Other nominal attributes are not supported by the Prolog notation.)<br/> |
|---|
| 66 | * <br/> |
|---|
| 67 | * |
|---|
| 68 | * @author Gabi Schmidberger (gabi@cs.waikato.ac.nz) |
|---|
| 69 | * @version $Revision: 1.5 $ |
|---|
| 70 | **/ |
|---|
| 71 | |
|---|
| 72 | public class Test |
|---|
| 73 | implements Serializable, RevisionHandler { |
|---|
| 74 | |
|---|
| 75 | /** for serialization */ |
|---|
| 76 | static final long serialVersionUID = -8890645875887157782L; |
|---|
| 77 | |
|---|
| 78 | /** the attribute index */ |
|---|
| 79 | int m_AttIndex; |
|---|
| 80 | |
|---|
| 81 | /** the split */ |
|---|
| 82 | double m_Split; |
|---|
| 83 | |
|---|
| 84 | /** whether to negate the test */ |
|---|
| 85 | boolean m_Not; |
|---|
| 86 | |
|---|
| 87 | /** the dataset */ |
|---|
| 88 | Instances m_Dataset; |
|---|
| 89 | |
|---|
| 90 | /** |
|---|
| 91 | * Constructor |
|---|
| 92 | * |
|---|
| 93 | * @param i the attribute index |
|---|
| 94 | * @param s the split |
|---|
| 95 | * @param dataset the dataset |
|---|
| 96 | */ |
|---|
| 97 | public Test(int i, double s, Instances dataset) { |
|---|
| 98 | m_AttIndex = i; |
|---|
| 99 | m_Split = s; |
|---|
| 100 | m_Dataset = dataset; |
|---|
| 101 | |
|---|
| 102 | m_Not = false; |
|---|
| 103 | } |
|---|
| 104 | |
|---|
| 105 | /** |
|---|
| 106 | * Constructor |
|---|
| 107 | * |
|---|
| 108 | * @param i the attribute index |
|---|
| 109 | * @param s the split |
|---|
| 110 | * @param dataset the dataset |
|---|
| 111 | * @param n whether to negate the test |
|---|
| 112 | */ |
|---|
| 113 | public Test(int i, double s, Instances dataset, boolean n) { |
|---|
| 114 | m_AttIndex = i; |
|---|
| 115 | m_Split = s; |
|---|
| 116 | m_Dataset = dataset; |
|---|
| 117 | m_Not = n; |
|---|
| 118 | } |
|---|
| 119 | |
|---|
| 120 | /** |
|---|
| 121 | * Negates the test. |
|---|
| 122 | * |
|---|
| 123 | * @return the test itself negated |
|---|
| 124 | */ |
|---|
| 125 | public Test getNot() { // returns a modified copy |
|---|
| 126 | return new Test(m_AttIndex, m_Split, m_Dataset, m_Not ? false : true); |
|---|
| 127 | } |
|---|
| 128 | |
|---|
| 129 | /** |
|---|
| 130 | * Determines whether an instance passes the test. |
|---|
| 131 | * |
|---|
| 132 | * @param inst the instance |
|---|
| 133 | * @return true if the instance satisfies the test, false otherwise |
|---|
| 134 | * @throws Exception if something goes wrong |
|---|
| 135 | */ |
|---|
| 136 | public boolean passesTest(Instance inst) throws Exception { |
|---|
| 137 | if (inst.isMissing(m_AttIndex)) return false; // missing values fail |
|---|
| 138 | |
|---|
| 139 | boolean isNominal = inst.attribute(m_AttIndex).isNominal(); |
|---|
| 140 | double attribVal = inst.value(m_AttIndex); |
|---|
| 141 | if (!m_Not) { |
|---|
| 142 | if (isNominal) { |
|---|
| 143 | if (((int) attribVal) != ((int) m_Split)) return false; |
|---|
| 144 | } |
|---|
| 145 | else if (attribVal >= m_Split) return false; |
|---|
| 146 | } else { |
|---|
| 147 | if (isNominal) { |
|---|
| 148 | if (((int) attribVal) == ((int) m_Split)) return false; |
|---|
| 149 | } |
|---|
| 150 | else if (attribVal < m_Split) return false; |
|---|
| 151 | } |
|---|
| 152 | return true; |
|---|
| 153 | } |
|---|
| 154 | |
|---|
| 155 | /** |
|---|
| 156 | * Returns the test represented by a string. |
|---|
| 157 | * |
|---|
| 158 | * @return a string representing the test |
|---|
| 159 | */ |
|---|
| 160 | public String toString() { |
|---|
| 161 | return (m_Dataset.attribute(m_AttIndex).name() + " " + |
|---|
| 162 | testComparisonString()); |
|---|
| 163 | } |
|---|
| 164 | |
|---|
| 165 | /** |
|---|
| 166 | * Returns the test represented by a string in Prolog notation. |
|---|
| 167 | * |
|---|
| 168 | * @return a string representing the test in Prolog notation |
|---|
| 169 | */ |
|---|
| 170 | public String toPrologString() { |
|---|
| 171 | Attribute att = m_Dataset.attribute(m_AttIndex); |
|---|
| 172 | StringBuffer str = new StringBuffer(); |
|---|
| 173 | String attName = m_Dataset.attribute(m_AttIndex).name(); |
|---|
| 174 | if (att.isNumeric()) { |
|---|
| 175 | str = str.append(attName + " "); |
|---|
| 176 | if (m_Not) str = str.append(">= " + Utils.doubleToString(m_Split, 3)); |
|---|
| 177 | else str = str.append("< " + Utils.doubleToString(m_Split, 3)); |
|---|
| 178 | } else { |
|---|
| 179 | String value = att.value((int)m_Split); |
|---|
| 180 | |
|---|
| 181 | if (value == "false") { str = str.append("not(" + attName + ")"); } |
|---|
| 182 | else { str = str.append(attName); } |
|---|
| 183 | } |
|---|
| 184 | return str.toString(); |
|---|
| 185 | } |
|---|
| 186 | |
|---|
| 187 | /** |
|---|
| 188 | * Gives a string representation of the test, starting from the comparison |
|---|
| 189 | * symbol. |
|---|
| 190 | * |
|---|
| 191 | * @return a string representing the test |
|---|
| 192 | */ |
|---|
| 193 | private String testComparisonString() { |
|---|
| 194 | Attribute att = m_Dataset.attribute(m_AttIndex); |
|---|
| 195 | if (att.isNumeric()) { |
|---|
| 196 | return ((m_Not ? ">= " : "< ") + Utils.doubleToString(m_Split,3)); |
|---|
| 197 | } |
|---|
| 198 | else { |
|---|
| 199 | if (att.numValues() != 2) |
|---|
| 200 | return ((m_Not ? "!= " : "= ") + att.value((int)m_Split)); |
|---|
| 201 | else return ("= " |
|---|
| 202 | + (m_Not ? |
|---|
| 203 | att.value((int)m_Split == 0 ? 1 : 0) : att.value((int)m_Split))); |
|---|
| 204 | } |
|---|
| 205 | } |
|---|
| 206 | |
|---|
| 207 | /** |
|---|
| 208 | * Gives a string representation of the test in Prolog notation, starting |
|---|
| 209 | * from the comparison symbol. |
|---|
| 210 | * |
|---|
| 211 | * @return a string representing the test in Prolog notation |
|---|
| 212 | */ |
|---|
| 213 | private String testPrologComparisonString() { |
|---|
| 214 | Attribute att = m_Dataset.attribute(m_AttIndex); |
|---|
| 215 | if (att.isNumeric()) { |
|---|
| 216 | return ((m_Not ? ">= " : "< ") + Utils.doubleToString(m_Split,3)); |
|---|
| 217 | } |
|---|
| 218 | else { |
|---|
| 219 | if (att.numValues() != 2) |
|---|
| 220 | return ((m_Not ? "!= " : "= ") + att.value((int)m_Split)); |
|---|
| 221 | else return ("= " |
|---|
| 222 | + (m_Not ? att.value((int)m_Split == 0 ? 1 : 0) |
|---|
| 223 | : att.value((int)m_Split))); |
|---|
| 224 | } |
|---|
| 225 | } |
|---|
| 226 | |
|---|
| 227 | /** |
|---|
| 228 | * Compares the test with the test that is given as parameter. |
|---|
| 229 | * |
|---|
| 230 | * @param t the test the object is compared to |
|---|
| 231 | * @return true if the two Tests are equal |
|---|
| 232 | */ |
|---|
| 233 | public boolean equalTo(Test t) { |
|---|
| 234 | return (m_AttIndex == t.m_AttIndex && m_Split == t.m_Split && m_Not == t.m_Not); |
|---|
| 235 | } |
|---|
| 236 | |
|---|
| 237 | /** |
|---|
| 238 | * Returns the revision string. |
|---|
| 239 | * |
|---|
| 240 | * @return the revision |
|---|
| 241 | */ |
|---|
| 242 | public String getRevision() { |
|---|
| 243 | return RevisionUtils.extract("$Revision: 1.5 $"); |
|---|
| 244 | } |
|---|
| 245 | } |
|---|
| 246 | |
|---|