Package adams.data.utils
Class SAXUtils
- java.lang.Object
-
- adams.data.utils.SAXUtils
-
public class SAXUtils extends Object
A helper class for SAX.- Version:
- $Revision$
- Author:
- dale (dale at waikato dot ac dot nz)
-
-
Constructor Summary
Constructors Constructor Description SAXUtils()
-
Method Summary
All Methods Static Methods Concrete Methods Modifier and Type Method Description static double[]
calcBreakPoints(int bins)
Calculate the break points for equal-frequency bins for a gaussian.static double[][]
calcDistMatrix(double[] bps)
Calculate the distance matrix for use in the MINDIST function.static double
minDist(double[] q, double[] c, double[][] distMatrix, int orig_n)
Calculate the distance between 2 SAX vectors.static double[]
PAA(double[] inarray, int numwindows)
Piecewise Aggregate Approximation.static double[]
toSAX(double[] inarray, int numwindows, double[] bps)
Convert a row in original space into SAX labels.
-
-
-
Method Detail
-
PAA
public static double[] PAA(double[] inarray, int numwindows)
Piecewise Aggregate Approximation.- Parameters:
inarray
- input arraynumwindows
- number of pieces- Returns:
- PAA
-
toSAX
public static double[] toSAX(double[] inarray, int numwindows, double[] bps)
Convert a row in original space into SAX labels. Assumes original space has been x normalised (rownorm?).- Parameters:
inarray
-numwindows
-bps
-- Returns:
-
calcDistMatrix
public static double[][] calcDistMatrix(double[] bps)
Calculate the distance matrix for use in the MINDIST function.- Parameters:
bps
- breakpoints- Returns:
- distance matrix
-
minDist
public static double minDist(double[] q, double[] c, double[][] distMatrix, int orig_n)
Calculate the distance between 2 SAX vectors.- Parameters:
q
- vector 1c
- vector 2distMatrix
- distance matrixorig_n
- original vector length- Returns:
- distance measure
-
calcBreakPoints
public static double[] calcBreakPoints(int bins)
Calculate the break points for equal-frequency bins for a gaussian.- Parameters:
bins
- number of bins- Returns:
- break points. Z numbers.
-
-