4.4 Summary

The jsl.utilities.statistic package defines a lot of functionality. Here is a summary of some of the useful classes and interfaces.

  1. CollectorIfc defines a set of collect() methods for collecting data. The method is overridden to permit the collection of a wide variety of data type. The collect() method is designed to collect values and a weight associated with the value. This allows the collection of weighted statistics. AbstractCollector is an abstract base class for building concrete sub-classes.
  2. SaveDataIfc defines methods for saving the observed data to arrays.
  3. WeightedStatisticIfc defines statistics that are computed on weighted data values. WeightedStatistic is a concrete implementation of the interface.
  4. AbstractStatistic is an abstract base class for defining statistics. Sub-classes of AbstractStatistic compute summary statistics of some kind.
  5. Histogram defines a class to collect statistics and tabulate data into bins.
  6. Statistic is a concrete implementation of AbstractStatistic allowing for a multitude of summary statistics.
  7. BatchStatistic is also a concrete implementation of AbstractStatistic that provides for summarizing data via a batching process.
  8. IntegerFrequency tabulates integer values into a frequencies by observed values, similar to a histogram.
  9. StateFrequency facilitates defining labeled states and tabulating visitation and transition statistics.
  10. StatisticXY collects statistics on \((x,y)\) pairs computing statistics on the \(x\) and \(y\) values separately, as well as the covariance and correlation between the observations within a pair.

The most important class within the statistics package is probably the Statistic class. This class summarizes the observed data into summary statistics such as: minimum, maximum, average, variance, standard deviation, lag-1 correlation, and count. In addition, confidence intervals can be formed on the observations based on the student-t distribution. Finally, there are useful static methods for computing statistics on arrays and for estimating sample sizes. The reader is encourage to review the JSL documentation for all of the functionality, including the ability to write nicely printed statistical results.