Package org.apache.commons.math3.stat.correlation

Class StorelessBivariateCovariance

java.lang.Object
org.apache.commons.math3.stat.correlation.StorelessBivariateCovariance
class StorelessBivariateCovariance extends Object
Bivariate Covariance implementation that does not require input data to be stored in memory.

This class is based on a paper written by Philippe Pébay: Formulas for Robust, One-Pass Parallel Computation of Covariances and Arbitrary-Order Statistical Moments, 2008, Technical Report SAND2008-6212, Sandia National Laboratories. It computes the covariance for a pair of variables. Use StorelessCovariance to estimate an entire covariance matrix.

Note: This class is package private as it is only used internally in the StorelessCovariance class.

Since:
3.0

  • Field Details

    • meanX

      private double meanX
      the mean of variable x

    • meanY

      private double meanY
      the mean of variable y

    • n

      private double n
      number of observations

    • covarianceNumerator

      private double covarianceNumerator
      the running covariance estimate

    • biasCorrected

      private boolean biasCorrected
      flag for bias correction

  • Constructor Details

    • StorelessBivariateCovariance

      StorelessBivariateCovariance()
      Create an empty StorelessBivariateCovariance instance with bias correction.

    • StorelessBivariateCovariance

      StorelessBivariateCovariance(boolean biasCorrection)
      Create an empty StorelessBivariateCovariance instance.
      Parameters:
      biasCorrection - if true the covariance estimate is corrected for bias, i.e. n-1 in the denominator, otherwise there is no bias correction, i.e. n in the denominator.

  • Method Details

    • increment

      public void increment(double x, double y)
      Update the covariance estimation with a pair of variables (x, y).
      Parameters:
      x - the x value
      y - the y value

    • append

      public void append(StorelessBivariateCovariance cov)
      Appends another bivariate covariance calculation to this. After this operation, statistics returned should be close to what would have been obtained by by performing all of the increment(double, double) operations in cov directly on this.
      Parameters:
      cov - StorelessBivariateCovariance instance to append.

    • getN

      public double getN()
      Returns the number of observations.
      Returns:
      number of observations

    • getResult

      public double getResult() throws NumberIsTooSmallException
      Return the current covariance estimate.
      Returns:
      the current covariance
      Throws:
      NumberIsTooSmallException - if the number of observations is < 2