Facemorph.Stats
Class PLSReducer

java.lang.Object
  Facemorph.Stats.PLSReducer

All Implemented Interfaces:

IniHandler, DataReducer, Regressor, StatisticalModel

Direct Known Subclasses:

sPLSSVD, sPLSSVD_DWH, WeightedPLS

public class PLSReducer
extends java.lang.Object
implements DataReducer, Regressor

Nested Class Summary

Nested classes/interfaces inherited from interface Facemorph.Stats.DataReducer

DataReducer.DataReducerInformation

Nested classes/interfaces inherited from interface Facemorph.Stats.Regressor

Regressor.RegressorException

Field Summary

protected BigMat	B
protected double[]	c
protected BigMat	C
protected BigMat	E
protected double	epsilon
protected BigMat	F
protected int	maxComponents
protected double	maxIter
protected double[]	meanX
protected double[]	meanY
protected int	noComponents
protected double[]	p
protected BigMat	P
protected double[]	q
protected BigMat	R
protected double[]	stdX
protected double[]	stdY
protected double[]	t
protected BigMat	T
protected double	targetDependentVariance
protected double	targetIndependentVariance
protected double[]	u
protected BigMat	U
protected double[]	varX
protected double[]	varY
protected double[]	w
protected BigMat	W
protected BigMat	X
protected BigMat	Y

Constructor Summary

PLSReducer()

Method Summary

BigMat	apply(BigMat in) Apply the statistical model on the input data supplied in Matrix form
int	build()
void	center()
BigMat	getBasisMatrix() Outputs the basis matrix that is the result of PCA analysis
double	getEpsilon() Get epsilon, the very small error margin used the determine if the function has finished minimising.
int	getInputDimensions()
void	getModelInformation(ModelInformation mi) Queries the statistical model about its abilities.
void	getModelInformation(ModelResults result) Query the (previously built) model for information about the model's components
int	getOutputDimensions()
double	getSumSquaredResiduals()
double	getTargetDependentVariance()
double	getTargetIndpendentVariance()
void	initialise()
void	iterate()
void	iterate(BigMat Ew, BigMat Fw)
static void	main(java.lang.String[] args)
java.lang.StringBuffer	outputStatistics()
boolean	outputStatistics(java.io.File statsFile)
double[]	predict(BigMat values) Predict using the model the values of the dependent from a set of independents
protected void	print()
boolean	read(java.io.File f)
boolean	read(iniFile file) Read from the current position in an iniFile.
BigMat	reconstructDependent(BigMat in) Reconstruct (approximately) a high dimensional input given the low dimensional output
double[]	reconstructDependent(double[] in)
BigMat	reconstructIndependent(BigMat in) Reconstruct (approximately) a high dimensional input given the low dimensional output
double[]	reconstructIndependent(double[] in) Reconstruct (approximately) a high dimensional input given the low dimensional output
BigMat	reduceDependent(BigMat in) Perform dimensionality reduction on the BigMat
double[]	reduceDependent(double[] in)
BigMat	reduceIndependent(BigMat in) Perform dimensionality reduction on the BigMat (in) using a Least Squares approximation
double[]	reduceIndependent(double[] in) Perform dimensionality reduction on the vector (in) using a Least Squares approximation
void	setDependentData(BigMat data) The dependent part of the regression
void	setDependentData(double[] data) The dependent part of the regression
void	setEpsilon(double e) Set epsilon, the very small error margin used the determine if the function has finished minimising.
void	setIndependentData(BigMat data) The independent part of the regression
void	setOutputDimensions(int d) Set the maximum number of output dimensions in the model (pre-build, undefined if set after building some classes will alter the dimensionality of the reduction others will not) If d is greater than the number input variables, the model will truncate at the maximum number of variables.
void	setTargetDependentVariance(double var) Stop calculating components when the variance explained in the dependent variable is greater than var
void	setTargetIndependentVariance(double var) Stop calculating components when the variance explained in the independent variable is greater than var
void	setTargetKaiserGuttman()
boolean	train(boolean stats) Perform multi-linear regression using the Ordinary Least Squares method.
boolean	write(java.io.File s) Writes to the file specified
boolean	write(iniFile file, java.lang.String name) Writes this Template to file (via a PrintStream)
boolean	write(java.io.PrintStream out) Writes this Template to file (via a PrintStream)

Methods inherited from class java.lang.Object

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

Field Detail

epsilon

protected double epsilon

maxIter

protected double maxIter

targetDependentVariance

protected double targetDependentVariance

targetIndependentVariance

protected double targetIndependentVariance

X

protected BigMat X

Y

protected BigMat Y

u

protected double[] u

t

protected double[] t

w

protected double[] w

p

protected double[] p

c

protected double[] c

q

protected double[] q

varX

protected double[] varX

varY

protected double[] varY

meanX

protected double[] meanX

meanY

protected double[] meanY

stdX

protected double[] stdX

stdY

protected double[] stdY

T

protected BigMat T

P

protected BigMat P

U

protected BigMat U

W

protected BigMat W

C

protected BigMat C

B

protected BigMat B

R

protected BigMat R

E

protected BigMat E

F

protected BigMat F

noComponents

protected int noComponents

maxComponents

protected int maxComponents

Constructor Detail

PLSReducer

public PLSReducer()

Method Detail

setTargetDependentVariance

public void setTargetDependentVariance(double var)

Description copied from interface: DataReducer

Stop calculating components when the variance explained in the dependent variable is greater than var

Specified by:

setTargetDependentVariance in interface DataReducer

Parameters:

var - variance to explain

getTargetDependentVariance

public double getTargetDependentVariance()

Specified by:

getTargetDependentVariance in interface DataReducer

Returns:

the target variance to explain on the dependent variable (not neccessarily the amount actually calculated

setTargetIndependentVariance

public void setTargetIndependentVariance(double var)

Description copied from interface: DataReducer

Stop calculating components when the variance explained in the independent variable is greater than var

Specified by:

setTargetIndependentVariance in interface DataReducer

Parameters:

var - variance to explain

getTargetIndpendentVariance

public double getTargetIndpendentVariance()

Specified by:

getTargetIndpendentVariance in interface DataReducer

Returns:

the target variance to explain on the independent variable (not neccessarily the amount actually calculated

setEpsilon

public void setEpsilon(double e)

Description copied from interface: DataReducer

Set epsilon, the very small error margin used the determine if the function has finished minimising. i.e. if the change is less than e stop minimising

Specified by:

setEpsilon in interface DataReducer

Parameters:

e - error value

getEpsilon

public double getEpsilon()

Description copied from interface: DataReducer

Get epsilon, the very small error margin used the determine if the function has finished minimising. i.e. if the change is less than e stop minimising

Specified by:

getEpsilon in interface DataReducer

Returns:

setIndependentData

public void setIndependentData(BigMat data)

Description copied from interface: StatisticalModel

The independent part of the regression

Specified by:

setIndependentData in interface StatisticalModel

setDependentData

public void setDependentData(BigMat data)

Description copied from interface: StatisticalModel

The dependent part of the regression

Specified by:

setDependentData in interface StatisticalModel

setDependentData

public void setDependentData(double[] data)

Description copied from interface: StatisticalModel

The dependent part of the regression

Specified by:

setDependentData in interface StatisticalModel

train

public boolean train(boolean stats)

Description copied from interface: StatisticalModel

Perform multi-linear regression using the Ordinary Least Squares method. Solves y = XB + e using b = inv(X'X)X'y

Specified by:

train in interface StatisticalModel

Parameters:

stats - calculate values for statistical analysis

Returns:

true if the regression succeeds.

center

public void center()

initialise

public void initialise()

iterate

public void iterate(BigMat Ew,
                    BigMat Fw)

build

public int build()

iterate

public void iterate()

reduceIndependent

public BigMat reduceIndependent(BigMat in)

Perform dimensionality reduction on the BigMat (in) using a Least Squares approximation

Specified by:

reduceIndependent in interface DataReducer

Parameters:

in - BigMat to reduce

Returns:

a shorted vector mapped in PLS space

reduceIndependent

public double[] reduceIndependent(double[] in)

Perform dimensionality reduction on the vector (in) using a Least Squares approximation

Specified by:

reduceIndependent in interface DataReducer

Parameters:

in - Vector to reduce

Returns:

a shorted vector mapped in PLS space

reconstructIndependent

public BigMat reconstructIndependent(BigMat in)

Description copied from interface: DataReducer

Reconstruct (approximately) a high dimensional input given the low dimensional output

Specified by:

reconstructIndependent in interface DataReducer

Parameters:

in - low dimensional matrix

Returns:

high dimensional matrix

reconstructDependent

public double[] reconstructDependent(double[] in)

Specified by:

reconstructDependent in interface DataReducer

setOutputDimensions

public void setOutputDimensions(int d)

Description copied from interface: DataReducer

Set the maximum number of output dimensions in the model (pre-build, undefined if set after building some classes will alter the dimensionality of the reduction others will not) If d is greater than the number input variables, the model will truncate at the maximum number of variables.

Specified by:

setOutputDimensions in interface DataReducer

Parameters:

d - target number of dimensions.

getInputDimensions

public int getInputDimensions()

Specified by:

getInputDimensions in interface StatisticalModel

getOutputDimensions

public int getOutputDimensions()

Specified by:

getOutputDimensions in interface StatisticalModel

write

public boolean write(iniFile file,
                     java.lang.String name)

Description copied from interface: IniHandler

Writes this Template to file (via a PrintStream)

Specified by:

write in interface IniHandler

Parameters:

file - The output ini file to fill with class data

Returns:

if written successfully

read

public boolean read(iniFile file)

Description copied from interface: IniHandler

Read from the current position in an iniFile. assumes that the iniFile is already in the object's root block

Specified by:

read in interface IniHandler

Returns:

true if successfully read.

predict

public double[] predict(BigMat values)

Description copied from interface: Regressor

Predict using the model the values of the dependent from a set of independents

Specified by:

predict in interface Regressor

Parameters:

values - the independent values

Returns:

set of predicted values

getSumSquaredResiduals

public double getSumSquaredResiduals()

Specified by:

getSumSquaredResiduals in interface Regressor

Returns:

The Sum of Squared Residuals of the regression.

outputStatistics

public boolean outputStatistics(java.io.File statsFile)

Specified by:

outputStatistics in interface StatisticalModel

outputStatistics

public java.lang.StringBuffer outputStatistics()

Specified by:

outputStatistics in interface StatisticalModel

getBasisMatrix

public BigMat getBasisMatrix()

Outputs the basis matrix that is the result of PCA analysis

Returns:

the basis functions as a matrix

read

public boolean read(java.io.File f)
             throws java.io.FileNotFoundException

Specified by:

read in interface Regressor

Throws:

java.io.FileNotFoundException

write

public boolean write(java.io.File s)

Description copied from interface: Regressor

Writes to the file specified

Specified by:

write in interface Regressor

Parameters:

s - the name of the file to write to

Returns:

returns true if all is OK

write

public boolean write(java.io.PrintStream out)

Description copied from interface: Regressor

Writes this Template to file (via a PrintStream)

Specified by:

write in interface Regressor

Parameters:

out - The output PrintStream

Returns:

if written successfully

reconstructIndependent

public double[] reconstructIndependent(double[] in)

Description copied from interface: DataReducer

Reconstruct (approximately) a high dimensional input given the low dimensional output

Specified by:

reconstructIndependent in interface DataReducer

Parameters:

in - low dimensional vector

Returns:

high dimensional matrix

reduceDependent

public double[] reduceDependent(double[] in)

Specified by:

reduceDependent in interface DataReducer

reduceDependent

public BigMat reduceDependent(BigMat in)

Description copied from interface: DataReducer

Perform dimensionality reduction on the BigMat

Specified by:

reduceDependent in interface DataReducer

Parameters:

in - BigMat to reduce

Returns:

a shorted vector mapped in dimensionally reduced space

reconstructDependent

public BigMat reconstructDependent(BigMat in)

Description copied from interface: DataReducer

Reconstruct (approximately) a high dimensional input given the low dimensional output

Specified by:

reconstructDependent in interface DataReducer

Parameters:

in - low dimensional matrix

Returns:

high dimensional matrix

print

protected void print()

main

public static void main(java.lang.String[] args)
                 throws java.io.IOException

Throws:

java.io.IOException

getModelInformation

public void getModelInformation(ModelInformation mi)

Description copied from interface: StatisticalModel

Queries the statistical model about its abilities.

Specified by:

getModelInformation in interface StatisticalModel

apply

public BigMat apply(BigMat in)

Description copied from interface: StatisticalModel

Apply the statistical model on the input data supplied in Matrix form

Specified by:

apply in interface StatisticalModel

Parameters:

in - matrix containing values for modeling

Returns:

values computed by the model.

setTargetKaiserGuttman

public void setTargetKaiserGuttman()

Specified by:

setTargetKaiserGuttman in interface DataReducer

getModelInformation

public void getModelInformation(ModelResults result)

Description copied from interface: StatisticalModel

Query the (previously built) model for information about the model's components

Specified by:

getModelInformation in interface StatisticalModel

Parameters:

result - ModelResults object to fill with appropriate information