com.mockturtlesolutions.snifflib.util

Class NNCalc

java.lang.Object

com.mockturtlesolutions.snifflib.util.NNCalc

public class NNCalc
extends java.lang.Object

Class for determining nearest neighbors. There are various techniques for determining nearest neighbors. In the 1D case it is hard to do better than a sorted dataset and then use a sliding window of width K. For higher dimensions this is no longer possible and the calculation of all pairs distances becomes computationally costly. Other techniques can be used.

Field Summary

Fields

Modifier and Type	Field and Description
boolean	AssumeSorted
DblMatrix	Distances
int[]	Indices
int	K
java.lang.String	Method
DblMatrix	NearestDistances
DblMatrix[]	NearestXData
Qdecomp	QTree
DblMatrix[]	XData

Constructor Summary

Constructors

Constructor and Description
NNCalc(DblMatrix[] XData) DblMatrix constructor.

Method Summary

Methods

Modifier and Type	Method and Description
DblSort	getKNearest(DblMatrix[] refpt) Get the K-nearest neighbors about the Ith coordinate.
void	setK(int newK) Set the neighborhood size.
void	setMethod(java.lang.String method) Set the nearest-neighbor search algorithm.

Methods inherited from class java.lang.Object

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

Field Detail

AssumeSorted

public boolean AssumeSorted

Method

public java.lang.String Method

K

public int K

XData

public DblMatrix[] XData

QTree

public Qdecomp QTree

NearestXData

public DblMatrix[] NearestXData

NearestDistances

public DblMatrix NearestDistances

Indices

public int[] Indices

Distances

public DblMatrix Distances

Constructor Detail

NNCalc

public NNCalc(DblMatrix[] XData)

DblMatrix constructor.

Method Detail

setK

public void setK(int newK)

Set the neighborhood size.

setMethod

public void setMethod(java.lang.String method)

Set the nearest-neighbor search algorithm.

getKNearest

public DblSort getKNearest(DblMatrix[] refpt)

Get the K-nearest neighbors about the Ith coordinate.