001 /*
002 * Created on Jul 25, 2005
003 *
004 */
005 package aima.test.learningtest;
006
007 import java.util.ArrayList;
008 import java.util.List;
009
010 import junit.framework.TestCase;
011 import aima.learning.framework.DataSet;
012 import aima.learning.framework.DataSetFactory;
013 import aima.learning.inductive.DecisionTree;
014 import aima.learning.learners.DecisionTreeLearner;
015 import aima.util.Util;
016
017 /**
018 * @author Ravi Mohan
019 *
020 */
021 public class DecisionTreeTest extends TestCase {
022 private static final String YES = "Yes";
023
024 private static final String NO = "No";
025
026 public void testActualDecisionTreeClassifiesRestaurantDataSetCorrectly()
027 throws Exception {
028 DecisionTreeLearner learner = new DecisionTreeLearner(
029 createActualRestaurantDecisionTree(), "Unable to clasify");
030 int[] results = learner.test(DataSetFactory.getRestaurantDataSet());
031 assertEquals(12, results[0]);
032 assertEquals(0, results[1]);
033 }
034
035 public void testInducedDecisionTreeClassifiesRestaurantDataSetCorrectly()
036 throws Exception {
037 DecisionTreeLearner learner = new DecisionTreeLearner(
038 createInducedRestaurantDecisionTree(), "Unable to clasify");
039 int[] results = learner.test(DataSetFactory.getRestaurantDataSet());
040 assertEquals(12, results[0]);
041 assertEquals(0, results[1]);
042 }
043
044 public void testStumpCreationForSpecifiedAttributeValuePair()
045 throws Exception {
046 DataSet ds = DataSetFactory.getRestaurantDataSet();
047 List<String> unmatchedValues = new ArrayList<String>();
048 unmatchedValues.add(NO);
049 DecisionTree dt = DecisionTree.getStumpFor(ds, "alternate", YES, YES,
050 unmatchedValues, NO);
051 assertNotNull(dt);
052 }
053
054 public void testStumpCreationForDataSet() throws Exception {
055 DataSet ds = DataSetFactory.getRestaurantDataSet();
056 List<DecisionTree> dt = DecisionTree.getStumpsFor(ds, YES,
057 "Unable to classify");
058 assertEquals(26, dt.size());
059 }
060
061 public void testStumpPredictionForDataSet() throws Exception {
062 DataSet ds = DataSetFactory.getRestaurantDataSet();
063 List<DecisionTree> trees = DecisionTree.getStumpsFor(ds, YES,
064 "Unable to classify");
065 // for (DecisionTree tree : trees){
066 // DecisionTreeLearner learner = new DecisionTreeLearner(tree,"Unable to
067 // Classify");
068 // int[] result = learner.test(ds);
069 // System.out.println("On stump " +tree.getAttributeName()+ " " +
070 // result[0]+ " successes "+ result[1]+ " failures");
071 // }
072 List<String> unmatchedValues = new ArrayList<String>();
073 unmatchedValues.add(NO);
074 DecisionTree tree = DecisionTree.getStumpFor(ds, "hungry", YES, YES,
075 unmatchedValues, "Unable to Classify");
076 DecisionTreeLearner learner = new DecisionTreeLearner(tree,
077 "Unable to Classify");
078 int[] result = learner.test(ds);
079 assertEquals(5, result[0]);
080 assertEquals(7, result[1]);
081 // System.out.println("On stump " +tree.getAttributeName()+ " " +
082 // result[0]+ " successes "+ result[1]+ " failures");
083
084 }
085
086 private static DecisionTree createInducedRestaurantDecisionTree() {// from
087 // AIMA
088 // 2nd
089 // ED
090 // Fig
091 // 18.6
092 // friday saturday node
093 DecisionTree frisat = new DecisionTree("fri/sat");
094 frisat.addLeaf(Util.YES, Util.YES);
095 frisat.addLeaf(Util.NO, Util.NO);
096
097 // type node
098 DecisionTree type = new DecisionTree("type");
099 type.addLeaf("French", Util.YES);
100 type.addLeaf("Italian", Util.NO);
101 type.addNode("Thai", frisat);
102 type.addLeaf("Burger", Util.YES);
103
104 // hungry node
105 DecisionTree hungry = new DecisionTree("hungry");
106 hungry.addLeaf(Util.NO, Util.NO);
107 hungry.addNode(Util.YES, type);
108
109 // patrons node
110 DecisionTree patrons = new DecisionTree("patrons");
111 patrons.addLeaf("None", Util.NO);
112 patrons.addLeaf("Some", Util.YES);
113 patrons.addNode("Full", hungry);
114
115 return patrons;
116
117 }
118
119 private static DecisionTree createActualRestaurantDecisionTree() {// from
120 // AIMA
121 // 2nd
122 // ED
123 // Fig
124 // 18.2
125
126 // raining node
127 DecisionTree raining = new DecisionTree("raining");
128 raining.addLeaf(Util.YES, Util.YES);
129 raining.addLeaf(Util.NO, Util.NO);
130
131 // bar node
132 DecisionTree bar = new DecisionTree("bar");
133 bar.addLeaf(Util.YES, Util.YES);
134 bar.addLeaf(Util.NO, Util.NO);
135
136 // friday saturday node
137 DecisionTree frisat = new DecisionTree("fri/sat");
138 frisat.addLeaf(Util.YES, Util.YES);
139 frisat.addLeaf(Util.NO, Util.NO);
140
141 // second alternate node to the right of the diagram below hungry
142 DecisionTree alternate2 = new DecisionTree("alternate");
143 alternate2.addNode(Util.YES, raining);
144 alternate2.addLeaf(Util.NO, Util.YES);
145
146 // reservation node
147 DecisionTree reservation = new DecisionTree("reservation");
148 frisat.addNode(Util.NO, bar);
149 frisat.addLeaf(Util.YES, Util.YES);
150
151 // first alternate node to the left of the diagram below waitestimate
152 DecisionTree alternate1 = new DecisionTree("alternate");
153 alternate1.addNode(Util.NO, reservation);
154 alternate1.addNode(Util.YES, frisat);
155
156 // hungry node
157 DecisionTree hungry = new DecisionTree("hungry");
158 hungry.addLeaf(Util.NO, Util.YES);
159 hungry.addNode(Util.YES, alternate2);
160
161 // wait estimate node
162 DecisionTree waitEstimate = new DecisionTree("wait_estimate");
163 waitEstimate.addLeaf(">60", Util.NO);
164 waitEstimate.addNode("30-60", alternate1);
165 waitEstimate.addNode("10-30", hungry);
166 waitEstimate.addLeaf("0-10", Util.YES);
167
168 // patrons node
169 DecisionTree patrons = new DecisionTree("patrons");
170 patrons.addLeaf("None", Util.NO);
171 patrons.addLeaf("Some", Util.YES);
172 patrons.addNode("Full", waitEstimate);
173
174 return patrons;
175
176 }
177 }