/*
* NBClassifierTest.java
* NetBeans JUnit based test
*
* Created on June 25, 2002, 3:59 PM
*/
package edu.msu.cme.rdp.classifier.train.validation;
import edu.msu.cme.rdp.classifier.train.LineageSequence;
import edu.msu.cme.rdp.classifier.train.LineageSequenceParser;
import edu.msu.cme.rdp.readseq.utils.orientation.GoodWordIterator;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileReader;
import java.io.IOException;
import java.io.Reader;
import java.util.*;
import junit.framework.Test;
import junit.framework.TestCase;
import junit.framework.TestSuite;
/**
*
* @author wangqion
*/
public class NBClassifierTest extends TestCase {
public NBClassifierTest(java.lang.String testName) {
super(testName);
}
public static void main(java.lang.String[] args) {
junit.textui.TestRunner.run(suite());
}
/** Test of calculateProb method, of class Classification.NBClassifier. */
public void testAssignClass() throws FileNotFoundException, IOException {
System.out.println("testassignClass");
boolean useSeed = true;
int min_bootstrap_words = 20;
Reader taxReader = new FileReader(System.class.getResource("/test/classifier/testTaxon.txt").getFile());
TreeFactory factory = new TreeFactory(taxReader);
File infileReader = new File(System.class.getResource("/test/classifier/testNBClassifierSet.fasta").getFile());
LineageSequenceParser parser = new LineageSequenceParser(infileReader);
while (parser.hasNext()) {
factory.addSequence((LineageSequence) parser.next());
}
//after all the training set is being parsed, calculate the prior probability for all the words.
factory.calculateWordPrior();
HashMap<String, HierarchyTree> nodeMap = new HashMap<String, HierarchyTree>();
HierarchyTree root = factory.getRoot();
root.getNodeMap("GENUS", nodeMap);
ArrayList<HierarchyTree> nodeList = new ArrayList(nodeMap.values());
// test the first sequence
File queryReader = new File(System.class.getResource("/test/classifier/testNBClassifierSet.fasta").getFile());
parser = new LineageSequenceParser(queryReader);
LineageSequence pSeq = parser.next();
GoodWordIterator iterator = new GoodWordIterator(pSeq.getSeqString());
NBClassifier classifier = new NBClassifier(factory, iterator.getWordArr(), nodeList, useSeed, min_bootstrap_words);
assertEquals(pSeq.getSeqName(), "XG1_child1");
ValidationClassificationResult result = classifier.assignClass();
assertEquals("G1", ((HierarchyTree) result.getBestClass()).getName());
assertTrue(0.1 > result.getPosteriorProb());
pSeq = parser.next();
// test the 3rd getSequence()
pSeq = parser.next();
assertEquals(pSeq.getSeqName(), "XG2_child1");
iterator = new GoodWordIterator(pSeq.getSeqString());
classifier = new NBClassifier(factory, iterator.getWordArr(), nodeList, useSeed, min_bootstrap_words);
result = classifier.assignClass();
assertEquals("G2", ((HierarchyTree) result.getBestClass()).getName());
assertTrue(0.2 > result.getPosteriorProb());
pSeq = parser.next();
// test the 5th getSequence()
pSeq = parser.next();
assertEquals(pSeq.getSeqName(), "XPh2G6_child1");
iterator = new GoodWordIterator(pSeq.getSeqString());
classifier = new NBClassifier(factory, iterator.getWordArr(), nodeList, useSeed, min_bootstrap_words);
result = classifier.assignClass();
assertEquals("G1", ((HierarchyTree) result.getBestClass()).getName());
assertTrue(0.2 > result.getPosteriorProb());
//test the 8th sequence in G7, it is the same as the 9th sequence in G8
// the classifier should randomly choose a genus (either G7 or G8) because the score will be tie
parser.next();
parser.next();
pSeq = parser.next();
assertEquals(pSeq.getSeqName(), "XPh2G7_child1");
iterator = new GoodWordIterator(pSeq.getSeqString());
int G7_count = 0;
int G8_count = 0;
for ( int run = 0; run < DecisionMaker.NUM_OF_RUNS; run++){
classifier = new NBClassifier(factory, iterator.getWordArr(), nodeList, useSeed, min_bootstrap_words);
result = classifier.assignClass();
if ( ((HierarchyTree) result.getBestClass()).getName().equals("G7")){
G7_count ++;
}else if ( ((HierarchyTree) result.getBestClass()).getName().equals("G8")){
G8_count ++;
}
}
assertEquals( G7_count + G8_count, DecisionMaker.NUM_OF_RUNS, 0.1);
// each genus should be choosen at least once
assertTrue(G7_count > 1);
assertTrue(G8_count > 1);
}
public static Test suite() {
TestSuite suite = new TestSuite(NBClassifierTest.class);
return suite;
}
}