composite

Type Members

1. class AggregationTimeSeries[T, U] extends TimeSeries[U]

   

   This time series generator is an aggregation of other generators.

2. case class ConditionalTimeSeries[T](condition: TimeSeries[Boolean], success: TimeSeries[T], failure: TimeSeries[T]) extends TimeSeries[T] with Product with Serializable

   

   A time series based on an underlying time series.

   A time series based on an underlying time series. The values of the underlying time series are forwarded by this time series iff a binary value from an other time series is true.

   If the value of the underlying time series is not defined, then the value of this time series is not defined. If the value of the binary time series is not defined, then the value of this time series is not defined.

   condition

   the binary time series used for determining if the values of the underlying time series must be forwarded.

   success

   the time series used when the condition is verified.

   failure

   the time series used when the condition is not verified.

3. case class CorrelatedTimeSeries(base: TimeSeries[Double], seed: Int, rho: Double) extends TimeSeries[Double] with Product with Serializable

   

   This time series is built as a correlation of an other time series.

   This time series is built as a correlation of an other time series.

   See http://www.sitmo.com/article/generating-correlated-random-numbers/ for explainations.

   base

   the time series generator on which this generator is based.

   seed

   the value used as seed for the random number generator. For a fixed seed and a fixed time series, the correlated values are deterministically generated.

   rho

   the correlation coefficient determining the *strongness* of the correlation. Must be in [0, 1]

4. class DivideTimeSeries extends BinaryTimeSeries[Double, Double]

   

   This time series generator divides a time series by an other one.

5. case class FunctionTimeSeries[T](generator: TimeSeries[T], f: (LocalDateTime, T) ⇒ Option[T]) extends TimeSeries[T] with Product with Serializable

   

   A time series in which each value is defined as the function of the corresponding value of an other time series.

   A time series in which each value is defined as the function of the corresponding value of an other time series.

   generator

   the underlying generator

   f

   the function to apply to defined values

6. case class SlidingWindowTimeSeries[T](base: TimeSeries[T], duration: Duration, n: Int, aggregator: (Seq[(Duration, T)]) ⇒ Option[T]) extends IndependantTimeSeries[T] with Product with Serializable

   

   A time series that aggregates recent values of an other time series.

   A time series that aggregates recent values of an other time series. A typical use case is a mobile average time series.

   If no values are available for a given time window, a None value is retrieved.

   The number of periods in the sliding windows will only rely on the call frequency of this time series, for the underlying time series will only provide values for the times specified by this time series.

   base

   the underlying time series on which this time series is based.

   duration

   the time in the past this time series must consider for computing the sliding window. Values that relate to times before this will be ignored.

   n

   the number of points to consider in the time period.

   aggregator

   the function used to aggregate values.

7. case class TimeShiftTimeSeries[T](generator: TimeSeries[T], shift: Duration) extends TimeSeries[T] with Product with Serializable

   

   A time series based on an other time series, and for which time is shifted.

   A time series based on an other time series, and for which time is shifted.

   generator

   The base time series.

   shift

   The time shift to apply, so that this.compute(t) == generator.compute(t+shift);

8. case class TransitionTimeSeries[T](first: TimeSeries[T], second: TimeSeries[T], time: LocalDateTime, transition: Option[(Duration, (T, T, Double) ⇒ T)]) extends TimeSeries[T] with Product with Serializable

   

   A transition time series takes its values from two successive time series: initially, a first time series is used for generating the desired values.

   A transition time series takes its values from two successive time series: initially, a first time series is used for generating the desired values. At a given time, a second time series is used instead of the first one.

   There may be a transition period during which values from both base time series are mixed in order to produce a new value.

   If a transition is specified, its effect begins at the specified transition time.

   During the transition, if one of the base time series does not provide any value, the optional value generated by the other base time series is used as it. If both base time series don't generate any value, the transition time series does not generate any value.

   first

   the first base time series used to generate values.

   second

   the second base time series used to generate values.

   time

   the time at which the transition begins.

   transition

   the duration of the optional transition, as well as the function that describes how values from both base time series are mixed. The two first parameters of this function are the values of the first and the second base time series (respectively), and the third parameter is a value between 0 and 1 expression the status of the transition from the first base time series (0) to the second one (1). If no transition is specified, the new time series instantaneously prevails.