Class EpsilonStatisticsCalculator


  • public class EpsilonStatisticsCalculator
    extends Object
    Compute statistics: mean, confidence interval of estimated ACER values \(\epsilon_k(\eta_i)\).
    • Constructor Detail

      • EpsilonStatisticsCalculator

        public EpsilonStatisticsCalculator​(boolean isWeightedByPeakCount,
                                           double confidenceLevel)
        Create an instance with the weighting option and confidence interval.
        Parameters:
        isWeightedByPeakCount - whether or not weighted (by peak count) mean to be used
        confidenceLevel - the confidence level
    • Method Detail

      • compute

        public EmpiricalACERStatistics compute​(double[][] observations,
                                               double[][] acers,
                                               int kStepMemory)
        Compute the statistics.
        Parameters:
        kStepMemory - k in the k-step memory assumption
        observations - the data (each row for one period, row lengths can be different)
        acers - the estimated ACER's (each row for different levels in one period)
        Returns:
        the estimated results