public void setup(final EvolutionState state, final Parameter base) {
super.setup(state, base);
try {
System.out.println(
"Total docs for cat "
+ IndexInfoStaticG.getCatnumberAsString()
+ " "
+ IndexInfoStaticG.totalTrainDocsInCat
+ " Total test docs for cat "
+ IndexInfoStaticG.totalTestDocsInCat);
ImportantWords iw = new ImportantWords();
wordArrayPos = iw.getF1WordList(false, true);
wordArrayNeg = iw.getF1WordList(false, false);
} catch (IOException e) {
e.printStackTrace();
}
}
public class ClassifyAndOrNotGA extends Problem implements SimpleProblemForm {
private IndexSearcher searcher = IndexInfoStaticG.getIndexSearcher();
private float F1train = 0;
private String[] wordArrayPos, wordArrayNeg;
private BooleanQuery query;
public void setup(final EvolutionState state, final Parameter base) {
super.setup(state, base);
try {
System.out.println(
"Total docs for cat "
+ IndexInfoStaticG.getCatnumberAsString()
+ " "
+ IndexInfoStaticG.totalTrainDocsInCat
+ " Total test docs for cat "
+ IndexInfoStaticG.totalTestDocsInCat);
ImportantWords iw = new ImportantWords();
wordArrayPos = iw.getF1WordList(false, true);
wordArrayNeg = iw.getF1WordList(false, false);
} catch (IOException e) {
e.printStackTrace();
}
}
public void evaluate(
final EvolutionState state,
final Individual ind,
final int subpopulation,
final int threadnum) {
if (ind.evaluated) return;
GAFit fitness = (GAFit) ind.fitness;
IntegerVectorIndividual intVectorIndividual = (IntegerVectorIndividual) ind;
query = new BooleanQuery(true);
for (int i = 0; i < (intVectorIndividual.genome.length - 1); i = i + 2) {
if (intVectorIndividual.genome[i] < 0
|| intVectorIndividual.genome[i] >= wordArrayPos.length
|| intVectorIndividual.genome[i] >= wordArrayNeg.length
|| intVectorIndividual.genome[i + 1] < 0
|| intVectorIndividual.genome[i + 1] >= wordArrayPos.length
|| intVectorIndividual.genome[i + 1] >= wordArrayNeg.length) continue;
int wordInd = intVectorIndividual.genome[i + 1];
switch (intVectorIndividual.genome[i]) {
case 0:
query.add(
new TermQuery(new Term(IndexInfoStaticG.FIELD_CONTENTS, wordArrayPos[wordInd])),
BooleanClause.Occur.SHOULD);
break;
case 1:
query.add(
new TermQuery(new Term(IndexInfoStaticG.FIELD_CONTENTS, wordArrayPos[wordInd])),
BooleanClause.Occur.MUST);
break;
case 2:
query.add(
new TermQuery(new Term(IndexInfoStaticG.FIELD_CONTENTS, wordArrayNeg[wordInd])),
BooleanClause.Occur.MUST_NOT);
break;
default:
query.add(
new TermQuery(new Term(IndexInfoStaticG.FIELD_CONTENTS, wordArrayPos[wordInd])),
BooleanClause.Occur.SHOULD);
}
}
try {
TotalHitCountCollector collector = new TotalHitCountCollector();
searcher.search(query, IndexInfoStaticG.catTrainF, collector);
final int positiveMatch = collector.getTotalHits();
collector = new TotalHitCountCollector();
searcher.search(query, IndexInfoStaticG.othersTrainF, collector);
final int negativeMatch = collector.getTotalHits();
F1train =
ClassifyQuery.f1(positiveMatch, negativeMatch, IndexInfoStaticG.totalTrainDocsInCat);
fitness.setTrainValues(positiveMatch, negativeMatch);
fitness.setF1Train(F1train);
fitness.setQuery(query);
} catch (IOException e) {
e.printStackTrace();
}
float rawfitness = F1train;
((SimpleFitness) intVectorIndividual.fitness).setFitness(state, rawfitness, false);
ind.evaluated = true;
}
}
