Package qupath.opencv.ml.objects.features

Class Normalizer

java.lang.Object

qupath.opencv.ml.objects.features.Normalizer

public class Normalizer
extends Object

Class to help with simple feature normalization, by adding an offset and then multiplying by a scaling factor.

Author:

Pete Bankhead

Method Summary

Modifier and Type	Method	Description
static Normalizer	createNormalizer(double[] offsets, double[] scales, double missingValue)	Create a Normalizer with the specified parameters.
double	getMissingValue()	Return the value that will be output after normalization if the computed value is not finite.
double	getOffset(int ind)	Get the offset for the specified feature
double	getScale(int ind)	Get the scale factor for the specified feature
boolean	isIdentity()	Returns true if this normalizer does not actually do anything.
int	nFeatures()	The total number of features supported by this Normalizer
double	normalizeFeature(int idx, double originalValue)	Normalize a single feature.

Methods inherited from class java.lang.Object

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

Method Details

createNormalizer

public static Normalizer createNormalizer(double[] offsets,
 double[] scales,
 double missingValue)

Create a Normalizer with the specified parameters.

Parameters:

offsets - value to add to each feature

scales - value to multiply each feature, after applying the offset

missingValue - replacement value for non-finite features

Returns:

a Normalizer initialized accordingly

normalizeFeature

public double normalizeFeature(int idx,
 double originalValue)

Normalize a single feature.

Parameters:

idx - the index of the feature; this is required to identify the corresponding offset and scale

originalValue - the original value of the feature

Returns:

the normalized value of the feature

isIdentity

public boolean isIdentity()

Returns true if this normalizer does not actually do anything. This is the case if all offsets are zero and all scales are 1.

Returns:

getMissingValue

public double getMissingValue()

Return the value that will be output after normalization if the computed value is not finite.

Returns:

nFeatures

public int nFeatures()

The total number of features supported by this Normalizer

Returns:

getOffset

public double getOffset(int ind)

Get the offset for the specified feature

Parameters:

ind - index of the feature

Returns:

getScale

public double getScale(int ind)

Get the scale factor for the specified feature

Parameters:

ind - index of the feature

Returns: