Class ACERByCounting


  • public class ACERByCounting
    extends Object
    Estimate epsilons by counting conditional exceedances from the observations.

    The R equivalent function is acer::acer.estimate_acer.

    • Constructor Detail

      • ACERByCounting

        public ACERByCounting​(double[] barrierLevels,
                              int kStepMemory)
        Create an instance for estimating epsilon for each of the given barrier levels.
        Parameters:
        barrierLevels - the barrier levels \(\eta_i\)
        kStepMemory - value of k in the assumption of k-step memory
    • Method Detail

      • estimateForMultiPeriods

        public double[][] estimateForMultiPeriods​(double[][] observations)
        Estimate for multiple periods. This method is suitable for both stationary and non-stationary time series. Estimates for multiple periods can then be used to compute the sample mean, standard deviation, and confidence interval.
        Parameters:
        observations - data for multiple periods (one row for one period)
        Returns:
        estimated ACER values, each row contains estimates for one period
      • estimateForOnePeriod

        public double[] estimateForOnePeriod​(double[] observations)
        Estimate for a single period.
        Parameters:
        observations - data for one period
        Returns:
        estimated ACER values, the i-th element = \(\hat{\epsilon_k}(\eta_i)\)