"Quantile indexes must be between 0 and the scale, which is " + scale);
    }
  }

  private static double[] longsToDoubles(long[] longs) {
    int len = longs.length;
    double[] doubles = new double[len];
    for (int i = 0; i < len; i++) {
      doubles[i] = longs[i];
    }
    return doubles;
  }

  private static double[] intsToDoubles(int[] ints) {
    int len = ints.length;

    List<Short> shorts = Arrays.asList((short) 0, (short) 1, (short) 2);
    List<Integer> ints = Arrays.asList(0, 1, 2);
    List<Float> floats = Arrays.asList((float) 0, (float) 1, (float) 2);
    List<Long> longs = Arrays.asList((long) 0, (long) 1, (long) 2);
    List<Double> doubles = Arrays.asList((double) 0, (double) 1, (double) 2);

    assertThat(Doubles.toArray(bytes)).isEqualTo(array);
    assertThat(Doubles.toArray(shorts)).isEqualTo(array);

   *
   * @param values a series of values, which will be converted to {@code double} values (this may
   *     cause loss of precision for longs of magnitude over 2^53 (slightly over 9e15))
   */
  public static Stats of(long... values) {
    StatsAccumulator accumulator = new StatsAccumulator();
    accumulator.addAll(values);
    return accumulator.snapshot();
  }

  /**

   * {@code expectedInsertions} that was used when constructing the filter.
   *
   * @since 22.0
   */
  public long approximateElementCount() {
    long bitSize = bits.bitSize();
    long bitCount = bits.bitCount();

    /**
     * Each insertion is expected to reduce the # of clear bits by a factor of

   *
   * @param values a series of values, which will be converted to {@code double} values (this may
   *     cause loss of precision for longs of magnitude over 2^53 (slightly over 9e15))
   */
  public void addAll(long... values) {
    for (long value : values) {
      add(value);
    }
  }

  /**
   * Adds the given values to the dataset. The stream will be completely consumed by this method.

    // Iteratively multiply the longs as big as they can go.
    for (long num = startingNumber; num <= n; num++) {
      // Check to see if the floor(log2(num)) + 1 has changed.
      if ((num & nextPowerOfTwo) != 0) {
        nextPowerOfTwo <<= 1;
        bits++;
      }
      // Get rid of the 2s in num.
      int tz = Long.numberOfTrailingZeros(num);
      long normalizedNum = num >> tz;
      shift += tz;

    List<Short> shorts = Arrays.asList((short) 0, (short) 1, (short) 2);
    List<Integer> ints = Arrays.asList(0, 1, 2);
    List<Float> floats = Arrays.asList((float) 0, (float) 1, (float) 2);
    List<Long> longs = Arrays.asList((long) 0, (long) 1, (long) 2);
    List<Double> doubles = Arrays.asList((double) 0, (double) 1, (double) 2);

    assertThat(Floats.toArray(bytes)).isEqualTo(array);
    assertThat(Floats.toArray(shorts)).isEqualTo(array);

    List<Short> shorts = Arrays.asList((short) 0, (short) 1, (short) 2);
    List<Integer> ints = Arrays.asList(0, 1, 2);
    List<Float> floats = Arrays.asList((float) 0, (float) 1, (float) 2);
    List<Long> longs = Arrays.asList((long) 0, (long) 1, (long) 2);
    List<Double> doubles = Arrays.asList((double) 0, (double) 1, (double) 2);

    assertThat(Doubles.toArray(bytes)).isEqualTo(array);
    assertThat(Doubles.toArray(shorts)).isEqualTo(array);

    List<Short> shorts = Arrays.asList((short) 0, (short) 1, (short) 2);
    List<Integer> ints = Arrays.asList(0, 1, 2);
    List<Float> floats = Arrays.asList((float) 0, (float) 1, (float) 2);
    List<Long> longs = Arrays.asList((long) 0, (long) 1, (long) 2);
    List<Double> doubles = Arrays.asList((double) 0, (double) 1, (double) 2);

    assertThat(Floats.toArray(bytes)).isEqualTo(array);
    assertThat(Floats.toArray(shorts)).isEqualTo(array);

        .of(LONG_MANY_VALUES_SUM_OF_SQUARES_OF_DELTAS / LONG_MANY_VALUES_COUNT);
    assertThat(LARGE_LONG_VALUES_STATS.populationVariance())
        .isWithin(ALLOWED_ERROR * Long.MAX_VALUE * Long.MAX_VALUE)
        .of(LARGE_LONG_VALUES_POPULATION_VARIANCE);
  }

  public void testPopulationStandardDeviation() {
    assertThrows(