Class EmpiricalACEREstimation


  • public class EmpiricalACEREstimation
    extends Object
    This class estimates empirical ACER values from the given observations.

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

    • Constructor Detail

      • EmpiricalACEREstimation

        public EmpiricalACEREstimation()
        Create an instance with default values. That is,
         
         this(300, true, new EpsilonStatisticsCalculator(true, 0.95), 2);
         
         
      • EmpiricalACEREstimation

        public EmpiricalACEREstimation​(int nLevels,
                                       boolean usePeaksOnly,
                                       EpsilonStatisticsCalculator statisticsCalculator,
                                       int kStepMemory)
        Create an instance for counting empirical ACERs.
        Parameters:
        nLevels - number of barrier levels to be used for estimation (e.g., 300)
        usePeaksOnly - perform estimation on peaks only
        statisticsCalculator - the calculator for computing the mean and the confidence interval
        kStepMemory - k in k-step memory assumption (2 is enough in most cases)
    • Method Detail

      • estimate

        public EmpiricalACER estimate​(double[][] observations)
        Estimate epsilon (or ACERs) from the given observations (each row is for one period).
        Parameters:
        observations - the multi-period observations
        Returns:
        estimated epsilons