runningPercentile()

Calculate the running percentile of a Stream<BigDecimal>.

Implementation Notes

This implementation is suitable for Stream<BigDecimal>, for a version that takes a user-specified mapping function see runningPercentileBy(). By default, nulls are ignored and play no part in calculations, see treatNullAs() and treatNullAsZero() below for ways to change this behavior. The default MathContext for all calculations is MathContext.DECIMAL64, but this can be overridden (see withMathContext(), below).

Each emitted value is the percentile of every element seen so far, so the last value emitted is the percentile of the whole stream. Values that fall between two ranks are linearly interpolated (the R-7 method, NumPy’s default). The percentile is converted to BigDecimal from its decimal representation, so 99.9 is exactly 99.9. The rank and interpolation are calculated exactly, and only the result is rounded to the MathContext.

Every value seen is retained in sorted order, so memory use grows with the length of the stream. For long or unbounded streams, consider movingPercentile() instead.

Signatures

runningPercentile(double percentile)

percentile - The percentile to calculate, between 0 and 100, inclusive

Additional Methods

Method	Purpose
`treatNullAsZero()`	When encountering a `null` value in a stream, treat it as `BigDecimal.ZERO` instead. See example.
`treatNullAs(BigDecimal replacement)`	When encountering a `null` value in a stream, treat it as the given `replacement` value instead. See example.
`treatNullAsOne()`	When encountering a `null` value in a stream, treat it as `BigDecimal.ONE` instead.
`withMathContext(MathContext mathContext)`	Replace the `MathContext` used for all mathematical operations performed by this gatherer. See example.
`withOriginal()`	Include the original input value from the stream in addition to the calculated value in a `WithOriginal` record. See example.

Examples

Running 90th percentile

Stream
    .of("1.0", "2.0", "10.0", "2.0", "5.0")
    .map(BigDecimal::new)
    .gather(Gatherers4j.runningPercentile(90))
    .toList();

// [
//   BigDecimal("1.0")
//   BigDecimal("1.9")
//   BigDecimal("8.4")
//   BigDecimal("7.6")
//   BigDecimal("8.0")
// ]

Showing nulls are ignored by default

Stream
    .of(null, null, new BigDecimal("10.0"), new BigDecimal("2.0"), new BigDecimal("1.0"))
    .gather(Gatherers4j.runningPercentile(90))
    .toList();

// [
//   BigDecimal("10.0")
//   BigDecimal("9.2")
//   BigDecimal("8.4")
// ]

Treating null as zero

Stream
    .of(null, null, new BigDecimal("10.0"), new BigDecimal("2.0"))
    .gather(Gatherers4j.runningPercentile(90).treatNullAsZero())
    .toList();

// [
//   BigDecimal("0")
//   BigDecimal("0")
//   BigDecimal("8.0")
//   BigDecimal("7.6")
// ]

Replacing null with another `BigDecimal`

Stream
    .of(null, null, new BigDecimal("10.0"), new BigDecimal("2.0"))
    .gather(Gatherers4j.runningPercentile(90).treatNullAs(new BigDecimal("3.5")))
    .toList();

// [
//   BigDecimal("3.5")
//   BigDecimal("3.5")
//   BigDecimal("8.7")
//   BigDecimal("8.05")
// ]

Specifying a new `MathContext`

Stream
    .of("1.111", "2.222", "10.333", "2.444")
    .map(BigDecimal::new)
    .gather(Gatherers4j
        .runningPercentile(90)
        .withMathContext(new MathContext(3, RoundingMode.DOWN))
    )
    .toList();

// [
//   BigDecimal("1.11")
//   BigDecimal("2.10")
//   BigDecimal("8.68")
//   BigDecimal("7.94")
// ]

Emitting a record containing the original and calculated values

Stream
    .of("1.0", "2.0", "10.0", "2.0")
    .map(BigDecimal::new)
    .gather(Gatherers4j.runningPercentile(90).withOriginal())
    .toList();

// [
//   WithOriginal[original=1.0, calculated=1.0]
//   WithOriginal[original=2.0, calculated=1.9]
//   WithOriginal[original=10.0, calculated=8.4]
//   WithOriginal[original=2.0, calculated=7.6]
// ]