for (int i = 0; i < 0x1000; i++) {
      double setSize = rng.nextDouble();
      if (setSize < emptySetProportion) {
        sets[i] = ImmutableSet.of();
      } else if (setSize < emptySetProportion + singletonSetProportion) {
        sets[i] = ImmutableSet.of(PRESENT);
      } else {
        sets[i] = ImmutableSet.of(PRESENT, new Object());
      }

      if (rng.nextDouble() < hitRate) {
        queries[i] = PRESENT;
      } else {

   * result is chosen from those possibilities uniformly at random.
   *
   * <p>Zero is treated as having log2 == 0.
   */
  static double randomDouble(int maxExponent) {
    double result = RANDOM_SOURCE.nextDouble();
    result = Math.scalb(result, RANDOM_SOURCE.nextInt(maxExponent + 1));
    return RANDOM_SOURCE.nextBoolean() ? result : -result;
  }

  /** Returns a random integer between zero and {@code MAX_EXPONENT}. */

  private double[] dataset;

  @Override
  protected void setUp() {
    dataset = new double[DATASET_SIZE];
    for (int i = 0; i < DATASET_SIZE; i++) {
      dataset[i] = RNG.nextDouble();
    }
  }

  public void testSingleQuantile_median() {
    double referenceValue = REFERENCE_ALGORITHM.singleQuantile(1, 2, dataset.clone());
    for (QuantilesAlgorithm algorithm : NON_REFERENCE_ALGORITHMS) {

    System.out.println("hit rate: " + hit / req);
  }

  // for proper distributions later:
  // import JSci.maths.statistics.ProbabilityDistribution;
  // int key = (int) dist.inverse(random.nextDouble());

  private double[] dataset;

  @Override
  protected void setUp() {
    dataset = new double[DATASET_SIZE];
    for (int i = 0; i < DATASET_SIZE; i++) {
      dataset[i] = RNG.nextDouble();
    }
  }

  public void testSingleQuantile_median() {
    double referenceValue = REFERENCE_ALGORITHM.singleQuantile(1, 2, dataset.clone());
    for (QuantilesAlgorithm algorithm : NON_REFERENCE_ALGORITHMS) {

  void setUp() {
    Random rng = new Random();
    for (int i = 0; i < 0x100; i++) {
      datasets[i] = new double[datasetSize];
      for (int j = 0; j < datasetSize; j++) {
        datasets[i][j] = rng.nextDouble();
      }
    }
  }

  private double[] dataset(int i) {
    // We must test on a fresh clone of the dataset each time. Doing sorts and quickselects on a

  void setUp() {
    Random rng = new Random();
    for (int i = 0; i < 0x100; i++) {
      datasets[i] = new double[datasetSize];
      for (int j = 0; j < datasetSize; j++) {
        datasets[i][j] = rng.nextDouble();
      }
    }
  }

  private double[] dataset(int i) {
    // We must test on a fresh clone of the dataset each time. Doing sorts and quickselects on a

  @BeforeExperiment
  void setUp() {
    Random rng = new Random();
    for (int i = 0; i < 0x100; i++) {
      values[i] = new double[n];
      for (int j = 0; j < n; j++) {
        values[i][j] = rng.nextDouble();
      }
    }
  }

  @Benchmark
  int meanAndVariance(int reps) {
    int tmp = 0;
    for (int i = 0; i < reps; i++) {

    System.out.println("hit rate: " + hit / req);
  }

  // for proper distributions later:
  // import JSci.maths.statistics.ProbabilityDistribution;
  // int key = (int) dist.inverse(random.nextDouble());

   * result is chosen from those possibilities uniformly at random.
   *
   * <p>Zero is treated as having log2 == 0.
   */
  static double randomDouble(int maxExponent) {
    double result = RANDOM_SOURCE.nextDouble();
    result = Math.scalb(result, RANDOM_SOURCE.nextInt(maxExponent + 1));
    return RANDOM_SOURCE.nextBoolean() ? result : -result;
  }

  /** Returns a random integer between zero and {@code MAX_EXPONENT}. */