at.mug.iqm.commons.util.plot

Class AppEntropy

java.lang.Object

at.mug.iqm.commons.util.plot.AppEntropy

public class AppEntropy
extends Object

Pincus Approximate Entropy Pincus S.M., Approximate entropy as a measure of system complexity, Proc. Natl. Acad. Sci. USA, Vol.88, 2297-2301, 1991

Changes

• 2014 02 changed float to double for r

Since:

2012 10

Author:

Helmut Ahammer

Field Summary

Fields

Modifier and Type	Field and Description
private int	numbDataPoints

Constructor Summary

Constructors

Constructor and Description
AppEntropy()

Method Summary

Modifier and Type	Method and Description
double	calcAppEntropy(Vector<Double> data1D, int m, double r, int d) This method calculates the approximate entropy
private Vector<Double>	calcLogCorrelations(Vector<Integer> numberOfCorrelations, int m, int d) This method calculates logarithm of correlations
private Double	calcMean(Vector<Double> data1D) This method calculates the mean of a data series
private Vector<Vector<Double>>	calcNewSeries(Vector<Double> data1D, int m, int d) This method calculates new data series
private Vector<Integer>	calcNumberOfCorrelations(Vector<Vector<Double>> newDataSeries, int m, double distR) This method calculates the number of correlations
private double	calcStandardDeviation(Vector<Double> data1D) This method calculates the standard deviation of a data series
private double	calcSumOfCorrelation(Vector<Double> correlations, int m, int d) This method calculates the sum of correlations
private double	calcVariance(Vector<Double> data1D) This method calculates the variance of a data series

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

numbDataPoints

private int numbDataPoints

Constructor Detail

AppEntropy

public AppEntropy()

Method Detail

calcMean

private Double calcMean(Vector<Double> data1D)

This method calculates the mean of a data series

Parameters:

data1D -

Returns:

the mean

calcVariance

private double calcVariance(Vector<Double> data1D)

This method calculates the variance of a data series

Parameters:

data1D -

Returns:

the variance

calcStandardDeviation

private double calcStandardDeviation(Vector<Double> data1D)

This method calculates the standard deviation of a data series

Parameters:

data1D -

Returns:

the standard deviation

calcNewSeries

private Vector<Vector<Double>> calcNewSeries(Vector<Double> data1D,
                                             int m,
                                             int d)

This method calculates new data series

Parameters:

data1D - 1D data vector

m - : number of new calculated time series (m = 2, Pincus et al. 1994)

Returns:

Vector of Series (vectors)

calcNumberOfCorrelations

private Vector<Integer> calcNumberOfCorrelations(Vector<Vector<Double>> newDataSeries,
                                                 int m,
                                                 double distR)

This method calculates the number of correlations

Parameters:

newDataSeries - vector of 1D vectors

distR - distance in %of SD

Returns:

Vector (Number of Correlations)

calcLogCorrelations

private Vector<Double> calcLogCorrelations(Vector<Integer> numberOfCorrelations,
                                           int m,
                                           int d)

This method calculates logarithm of correlations

Parameters:

numberOfCorrelations - vector of mumbers

m - number of new calculated time series (m = 2, Pincus et al. 1994)

d - delay

Returns:

Vector correlations

calcSumOfCorrelation

private double calcSumOfCorrelation(Vector<Double> correlations,
                                    int m,
                                    int d)

This method calculates the sum of correlations

Parameters:

correlations - vector of correlations

m - number of newly calculated time series (m = 2, Pincus et al.1994)

d - delay

Returns:

sumOfCorrelations double

calcAppEntropy

public double calcAppEntropy(Vector<Double> data1D,
                             int m,
                             double r,
                             int d)

This method calculates the approximate entropy

Parameters:

data1D - 1D data vector

m - number of newly calculated time series (m = 2, Pincus et al.1994) m should not be greater than N/3 (N number of data points)!

r - maximal distance radius r (10%sd < r < 25%sd sd = standard deviation of time series, Pincus et al. 1994)

d - delay

Returns:

Approximate Entropy (single double value)