io.github.ericmedvet.jgea.problem.extraction.ExtractionFitness Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of jgea.problem Show documentation
Problem (benchmarks and templates) for jgea.
The newest version!
/*-
 * ========================LICENSE_START=================================
 * jgea-problem
 * %%
 * Copyright (C) 2018 - 2024 Eric Medvet
 * %%
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *      http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 * =========================LICENSE_END==================================
 */

package io.github.ericmedvet.jgea.problem.extraction;

import io.github.ericmedvet.jgea.core.representation.graph.finiteautomata.Extractor;
import io.github.ericmedvet.jgea.core.util.IntRange;
import java.util.*;
import java.util.function.Function;

public class ExtractionFitness implements Function, List> {

  private final Aggregator aggregator;

  public ExtractionFitness(List sequence, Set desiredExtractions, Metric... metrics) {
    aggregator = new Aggregator<>(sequence, desiredExtractions, metrics);
  }

  public enum Metric {
    ONE_MINUS_PREC,
    ONE_MINUS_REC,
    ONE_MINUS_FM,
    SYMBOL_FNR,
    SYMBOL_FPR,
    SYMBOL_ERROR,
    SYMBOL_WEIGHTED_ERROR
  }

  private static class Aggregator implements Function, List> {

    private final List sequence;
    private final Set desiredExtractions;
    private final List metrics;
    private final BitSet desiredExtractionMask;
    private final int positiveSymbols;

    public Aggregator(List sequence, Set desiredExtractions, Metric... metrics) {
      this.sequence = sequence;
      this.desiredExtractions = desiredExtractions;
      this.metrics = Arrays.asList(metrics);
      desiredExtractionMask = buildMask(desiredExtractions, sequence.size());
      positiveSymbols = desiredExtractionMask.cardinality();
    }

    @Override
    public List apply(Set extractions) {
      Map values = new EnumMap<>(Metric.class);
      if (metrics.contains(Metric.ONE_MINUS_FM)
          || metrics.contains(Metric.ONE_MINUS_PREC)
          || metrics.contains(Metric.ONE_MINUS_REC)) {
        // precision and recall
        Set correctExtractions = new LinkedHashSet<>(extractions);
        correctExtractions.retainAll(desiredExtractions);
        double recall = (double) correctExtractions.size() / (double) desiredExtractions.size();
        double precision = (double) correctExtractions.size() / (double) extractions.size();
        double fMeasure = 2d * precision * recall / (precision + recall);
        values.put(Metric.ONE_MINUS_PREC, 1 - precision);
        values.put(Metric.ONE_MINUS_REC, 1 - recall);
        values.put(Metric.ONE_MINUS_FM, 1 - fMeasure);
      }
      if (metrics.contains(Metric.SYMBOL_ERROR)
          || metrics.contains(Metric.SYMBOL_FNR)
          || metrics.contains(Metric.SYMBOL_FPR)
          || metrics.contains(Metric.SYMBOL_WEIGHTED_ERROR)) {
        BitSet extractionMask = buildMask(extractions, sequence.size());
        int extractedSymbols = extractionMask.cardinality();
        extractionMask.and(desiredExtractionMask);
        double truePositiveSymbols = extractionMask.cardinality();
        double falseNegativeSymbols = positiveSymbols - truePositiveSymbols;
        double falsePositiveSymbols = extractedSymbols - truePositiveSymbols;
        double trueNegativeChars = desiredExtractionMask.length()
            - falsePositiveSymbols
            - truePositiveSymbols
            - falseNegativeSymbols;
        values.put(Metric.SYMBOL_FPR, falsePositiveSymbols / (trueNegativeChars + falsePositiveSymbols));
        values.put(Metric.SYMBOL_FNR, falseNegativeSymbols / (truePositiveSymbols + falseNegativeSymbols));
        values.put(
            Metric.SYMBOL_ERROR, (falsePositiveSymbols + falseNegativeSymbols) / (double) sequence.size());
        values.put(
            Metric.SYMBOL_WEIGHTED_ERROR,
            (falsePositiveSymbols / (trueNegativeChars + falsePositiveSymbols)
                    + falseNegativeSymbols / (truePositiveSymbols + falseNegativeSymbols))
                / 2d);
      }
      List results = new ArrayList<>(metrics.size());
      for (Metric metric : metrics) {
        results.add(values.get(metric));
      }
      return results;
    }

    public Set getDesiredExtractions() {
      return desiredExtractions;
    }

    public List getMetrics() {
      return metrics;
    }

    public List getSequence() {
      return sequence;
    }

    private Set intersections(IntRange range, Set others) {
      Set intersections = new HashSet<>();
      for (IntRange other : others) {
        if (range.overlaps(other)) {
          intersections.add(range.intersection(other).orElseThrow());
        }
      }
      return intersections;
    }
  }

  private static BitSet buildMask(Set extractions, int size) {
    BitSet bitSet = new BitSet(size);
    extractions.forEach(r -> bitSet.set(r.min(), r.max()));
    return bitSet;
  }

  @Override
  public List apply(Extractor e) {
    return aggregator.apply(e.extractNonOverlapping(aggregator.sequence));
  }

  public ExtractionFitness changeMetrics(Metric... metrics) {
    return new ExtractionFitness<>(aggregator.sequence, aggregator.desiredExtractions, metrics);
  }

  public Set getDesiredExtractions() {
    return aggregator.getDesiredExtractions();
  }

  public List getMetrics() {
    return aggregator.getMetrics();
  }

  public List getSequence() {
    return aggregator.sequence;
  }
}