URANIE::Modeler::TGaussianCorrFunction Class Reference

Description of the class TGaussianCorrFunction. More...

#include <TGaussianCorrFunction.h>

Inheritance diagram for URANIE::Modeler::TGaussianCorrFunction:

Collaboration diagram for URANIE::Modeler::TGaussianCorrFunction:

Public Member Functions
Constructor and Destructor
	TGaussianCorrFunction (Int_t nbDim, Double_t *corrLength=NULL)
	Default constructor. More...
virtual	~TGaussianCorrFunction ()
	Default destructor. More...
Get and Set methods
virtual void	setParameters (Double_t *newParams)
	Sets the parameters of the gaussian correlation function. More...
void	setCorrLengths (Double_t *corrLength)
	Sets the correlation lengths. More...
Tools
virtual TCorrelationFunction *	clone ()
	Return a clone of the correlation function. More...
virtual void	eval (Int_t n1, Double_t *x1, Int_t n2, Double_t *x2, Double_t *C)
	Compute the correlation between x1 and x2. More...
Printing Log
virtual void	printLog (Option_t *option="")
Public Member Functions inherited from URANIE::Modeler::TCorrelationFunction
	TCorrelationFunction ()
	Default constructor. More...
virtual	~TCorrelationFunction ()
	Default destructor. More...
Int_t	getNParameters ()
	Return the number of parameters of the correlation function. More...
Int_t	getNDimensions ()
	Return the number of dimensions of the input space. More...
Int_t	getNCorrLengths ()
	Return the number of correlation lengths. More...
Double_t *	getParameters ()
	Returns an array containing the parameters of the function. More...
Double_t *	getCorrLengths ()
	Returns an array containing only the correlation lengths. More...
void	setLog ()
void	unsetLog ()
void	changeLog ()
Bool_t	getLog ()

Additional Inherited Members
Public Attributes inherited from URANIE::Modeler::TCorrelationFunction
Bool_t	_blog
	Boolean to decide if the log information is shown or not. More...
Double_t *	parameters
	array containing the values of the function parameters. Correlation lengths are stored last. More...
Double_t *	_lCorrLengths
	array containing only the correlation lengths. More...
Int_t	nbDimensions
	number of dimensions of the input space More...
Int_t	nbParameters
	total number of parameters More...
Int_t	nbCorrLengths
	number of correlation lengths More...

Detailed Description

Description of the class TGaussianCorrFunction.

This class defines the shape of the correlation between two sets of variables as a gaussian:

\[ c(x1,x2) = exp\left(-\frac{||x1-x2||^2}{lc^2}\right) \]

where \(lc\) is the correlation length.

Constructor & Destructor Documentation

TGaussianCorrFunction()

URANIE::Modeler::TGaussianCorrFunction::TGaussianCorrFunction	(	Int_t	nbDim,
		Double_t *	corrLength = NULL
	)

Default constructor.

Parameters

nbDim	(Int_t): number of dimension of the input variables
corrLength	(Double_t*): array of the correlation lengths for the function (size: nbDim). If NULL, the correlation lengths are initialised to 0.1. (default = NULL).

Referenced by ClassImp().

~TGaussianCorrFunction()

virtual URANIE::Modeler::TGaussianCorrFunction::~TGaussianCorrFunction ( )

virtual

Default destructor.

Referenced by ClassImp().

Member Function Documentation

clone()

virtual TCorrelationFunction* URANIE::Modeler::TGaussianCorrFunction::clone ( )

virtual

Return a clone of the correlation function.

The clone must be destroyed by the object (or user) calling the function.

Implements URANIE::Modeler::TCorrelationFunction.

Referenced by ClassImp().

eval()

virtual void URANIE::Modeler::TGaussianCorrFunction::eval ( Int_t  n1, Double_t *  x1, Int_t  n2, Double_t *  x2, Double_t *  C  )

virtual

Compute the correlation between x1 and x2.

The order of the indices in the flat arrays are the following: x[i+j*d] = X[i,j] where X is a matrix of size (d x n), i is the line index, and j the column index. The results are stored in the flat array C.

Warning

Be aware that the matrix corresponding to flat arrays x1 and x2 have input dimensions as LINE, not columns. This means that for a set of point {(a1,b1),(a2,b2),(a3,b3)}, the corresponding matrix X will be: \( \left( \begin{array}{ccc} a1 & a2 & a3 \\ b1 & b2 & b3 \end{array} \right) \)
and the corresponding flat array x = {a1,b1,a2,b2,a3,b3}.

Parameters

n1	(Int_t): number of elements of x1
x1	(Double_t*): data set x1 represented as a flat array (size: d * n1)
n2	(Int_t): number of elements of x2
x2	(Double_t*): data set x2 represented as a flat array (size: d * n2)
C	(Double_t*): correlation matrix represented as a flat array (size: n1 * n2)

Exceptions

UErrorExceptions

if the gpLib function returns 1 (evaluation failed)

Implements URANIE::Modeler::TCorrelationFunction.

Referenced by ClassImp().

printLog()

virtual void URANIE::Modeler::TGaussianCorrFunction::printLog ( Option_t *  option = "" )

virtual

Reimplemented from URANIE::Modeler::TCorrelationFunction.

Referenced by ClassImp().

setCorrLengths()

void URANIE::Modeler::TGaussianCorrFunction::setCorrLengths ( Double_t *  corrLength )

inlinevirtual

Sets the correlation lengths.

Parameters

corrLength

(Double_t*): array containing the correlation length parameter. If the number of dimension of the inputs is \(nbDim\), the array's size is \(nbDim\).

Implements URANIE::Modeler::TCorrelationFunction.

References setParameters().

setParameters()

virtual void URANIE::Modeler::TGaussianCorrFunction::setParameters ( Double_t *  newParams )

virtual

Sets the parameters of the gaussian correlation function.

Parameters

newParams

(Double_t*): array containing the parameters of the correlation function. If the number of dimension of the inputs is \(nbDim\), the array's size is \(nbDim\). It is supposed to be of the form: \[ newParams[nbDim] = {lc[0],..., lc[nbDim-1]} \] where \(lc\) is the correlation length parameter (array of size \(nbDim\)).

Implements URANIE::Modeler::TCorrelationFunction.

Referenced by ClassImp(), and setCorrLengths().