Description of the class TTester. More...

#include <TTester.h>

Inheritance diagram for URANIE::UncertModeler::TTester:

[legend]

Collaboration diagram for URANIE::UncertModeler::TTester:

Public Types
enum	ETestType { kKolmogorovSmirnov , kAndersonDarling , kCramerVonMises , kUnknown }

enum	ECase { kCase0 , kCase1 , kCase2 , kCase3 , kCaseUnknown }

Public Member Functions
Constructor and Destructor
	TTester (URANIE::DataServer::TDataServer tds, ETestType nTestType, const char varexpinput, Option_t *option="")
	Constructor by the type of Test.

virtual	~TTester ()
	Default constructor.

Setting and Getting attributes
Int_t	getID ()

Double_t	getScore ()
	Get the score of the last test.

Double_t	getModifiedScore ()
	Get the modified score of the last test.

Double_t	getPValue ()
	Get the pValue of the last test.

ECase	getCase ()
	getCaseOfGoF

Computes the score
Double_t	computeScore (Option_t *option="")
	Compute the score of the statistic.

virtual Double_t	computeTheCurrentScore (Int_t i, Double_t dzi)=0
	The true computation with a vector of information and return the current value.

virtual void	computeModifiedScore (URANIE::DataServer::TStochasticAttribute::ELawType tlaw)=0
	Compute the modified value of the statistic.

virtual const char *	getTitleStatistic ()=0
	Return the name of the statistic.

void	addUniformLaw (Double_t min, Double_t max, Color_t color=kBlack, TString stitle="")
	Add in the current graph the CDF of a uniform law defined by theses parameters $\min, \max$ .

void	addNormalLaw (Double_t mu, Double_t sigma, Color_t color=kBlack, TString stitle="")
	Add in the current graph the CDF of a normal law defined by theses parameters ${\cal N} ( \mu, \sigma )$ .

void	addLogNormalLaw (Double_t mu, Double_t sigma, Color_t color=kBlack, TString stitle="")
	Adds in the current graph the CDF of a LogNormal law defined by theses parameters $\mu, \sigma$ of the $ln(X) \sim {\cal N} ( \mu, \sigma )$ law.

void	addWeibullLaw (Double_t beta, Double_t m, Color_t color=kBlack, TString stitle="")
	Add in the current graph the CDF of a Weibull law defined by theses parameters ${\cal W}( lambda, k)$ .

void	addGammaLaw (Double_t alpha, Double_t beta, Color_t color=kBlack, TString stitle="")
	Adds in the current graph the CDF of a Gamma law defined by theses parameters ${\cal G}amma ( \alpha, \beta)$ .

void	addBetaLaw (Double_t alpha, Double_t beta, Double_t min, Double_t max, Color_t color=kBlack, TString stitle="")
	Add in the current graph the CDF of a Beta law defined by theses parameters $\alpha, \beta, \min, \max$ .

void	treatOption (Option_t *option)
	Treats the options of the computeScore method.

Computes the PValues
from the book D'Agostino and Stephens "Goodness-of-Fit Techniques" Depends on the case : Case0 : The parameters of the law are defined Case3 : The parameters of the law are not defined; then they are estimated by the MLE ("Maximum LikeLihood Estimate"). and the type of the EDF test (KS, CvM ou AD)
void	computePValue (URANIE::DataServer::TStochasticAttribute::ELawType tlaw)
	Compute the PValue of the test.

Double_t	computePValue_KS_CaseAll (Double_t dz)
	Compute the PValue of the test KS in all cases (Case0 && Case3)

Double_t	computePValue_AD_Case0 (Double_t dz)
	Compute the PValue of the test AD in case Case0 (all parameters of the PDF are given)

Double_t	computePValue_AD_Case3 (Double_t dz, URANIE::DataServer::TStochasticAttribute::ELawType tlaw)
	Compute the PValue of the test AD in case Case3 (all parameters of the PDF are estimated by MLE)

Double_t	computePValue_CvM_Case3 (Double_t dz, URANIE::DataServer::TStochasticAttribute::ELawType tlaw)
	Compute the PValue of the test CvM in case Case3 (all parameters of the PDF are estimated by MLE)

Log Printing
void	setLog ()

void	unsetLog ()

void	changeLog ()

Bool_t	getLog ()

virtual void	printLog (Option_t *option="")

Public Attributes
ETestType	_nTestType
	The type of test.

ECase	_nCase

URANIE::DataServer::TDataServer *	_tds
	Pointer to a TDS.

Bool_t	_blog
	Boolean for edit the log.

Bool_t	_bquiet
	Boolean for being quiet.

Protected Member Functions
Protected Methods
void	findParametersLaw (TString soption, TVectorD &vec)
	Finds the parameters in the option string.

Int_t	nTimesValueLTvalVector (Double_t dval, TVectorD vecValues)

TGraph *	createGraph (Color_t color=kBlack)

Protected Attributes
Long64_t	_nS
	the size of selected values

Double_t	_pValue
	The pValue of the last test.

Double_t	_dScore
	The score of the last test.

Double_t	_dModifiedScore
	The Modified value score of the last test.

TString	_sAttributes
	The string of attributes separated by the colon ":" character.

TList *	_listOfLaws
	The list of law to test.

Bool_t	_bDrawSame
	Boolean for drawing in the current pad.

Detailed Description

Description of the class TTester.

These class implements the tests of fit based on the Empirical Distribution Function ("EDF"). The EDF is a step function, calculated from the sample, wich estimates the population distribution.

See also: Goofness-of-fit techniques, R. D'Agostini and M. Stephens, Dekker

Member Enumeration Documentation

◆ ECase

enum URANIE::UncertModeler::TTester::ECase

Enumerator
kCase0
kCase1
kCase2
kCase3
kCaseUnknown

◆ ETestType

enum URANIE::UncertModeler::TTester::ETestType

Enumerator
kKolmogorovSmirnov
kAndersonDarling
kCramerVonMises
kUnknown

Constructor & Destructor Documentation

◆ TTester()

URANIE::UncertModeler::TTester::TTester	(	URANIE::DataServer::TDataServer *	tds,
		ETestType	nTestType,
		const char *	varexpinput,
		Option_t *	option = `""`
	)

Constructor by the type of Test.

Referenced by ClassImp().

◆ ~TTester()

virtual URANIE::UncertModeler::TTester::~TTester ( )

virtual

Default constructor.

Referenced by ClassImp().

Member Function Documentation

◆ addBetaLaw()

void URANIE::UncertModeler::TTester::addBetaLaw	(	Double_t	alpha,
		Double_t	beta,
		Double_t	min,
		Double_t	max,
		Color_t	color = `kBlack`,
		TString	stitle = `""`
	)

Add in the current graph the CDF of a Beta law defined by theses parameters $\alpha, \beta, \min, \max$ .

Parameters

alpha	(Double_t) the alpha value of the law
beta	(Double_t) the beta value of the law
min	(Double_t) the minnimum value of the law
max	(Double_t) the maximum value of the law
color	(Color_t)[kBlack] the color of the line of the CDF's law
stitle	(TString)[] the title of the entry in the legend box

Referenced by ClassImp().

◆ addGammaLaw()

void URANIE::UncertModeler::TTester::addGammaLaw	(	Double_t	alpha,
		Double_t	beta,
		Color_t	color = `kBlack`,
		TString	stitle = `""`
	)

Adds in the current graph the CDF of a Gamma law defined by theses parameters ${\cal G}amma ( \alpha, \beta)$ .

where the CDF formula is : $F(x) = inc_gamma( \alpha, x/scale )$

Parameters

alpha	(Double_t) The scale parameter $\alpha$ of the Gamma law
beta	(Double_t) The shape parameter $\beta$ of the the Gamma law
color	(Color_t)[kBlack] The color of the line of the CDF's law
stitle	(TString)[] The title of the entry in the legend box

Referenced by ClassImp().

◆ addLogNormalLaw()

void URANIE::UncertModeler::TTester::addLogNormalLaw	(	Double_t	mu,
		Double_t	sigma,
		Color_t	color = `kBlack`,
		TString	stitle = `""`
	)

Adds in the current graph the CDF of a LogNormal law defined by theses parameters $\mu, \sigma$ of the $ln(X) \sim {\cal N} ( \mu, \sigma )$ law.

Parameters

mu	(Double_t) The mean $\mu$ of the law
sigma	(Double_t) The standard-deviation $\sigma$ of the law
color	(Color_t)[kBlack] The color of the line of the CDF's law
stitle	(TString)[] The title of the entry in the legend box

Referenced by ClassImp().

◆ addNormalLaw()

void URANIE::UncertModeler::TTester::addNormalLaw	(	Double_t	mu,
		Double_t	sigma,
		Color_t	color = `kBlack`,
		TString	stitle = `""`
	)

Add in the current graph the CDF of a normal law defined by theses parameters ${\cal N} ( \mu, \sigma )$ .

Parameters

mu	(Double_t) the mean $\mu$ of the law
sigma	(Double_t) the standard-deviation $\sigma$ of the law
color	(Color_t)[kBlack] the color of the line of the CDF's law
stitle	(TString)[] the title of the entry in the legend box

Referenced by ClassImp().

◆ addUniformLaw()

void URANIE::UncertModeler::TTester::addUniformLaw	(	Double_t	min,
		Double_t	max,
		Color_t	color = `kBlack`,
		TString	stitle = `""`
	)

Add in the current graph the CDF of a uniform law defined by theses parameters $\min, \max$ .

Parameters

min	(Double_t) the minnimum value of the law
max	(Double_t) the maximum value of the law
color	(Color_t)[kBlack] the color of the line of the CDF's law
stitle	(TString)[] the title of the entry in the legend box

Referenced by ClassImp().

◆ addWeibullLaw()

void URANIE::UncertModeler::TTester::addWeibullLaw	(	Double_t	beta,
		Double_t	m,
		Color_t	color = `kBlack`,
		TString	stitle = `""`
	)

Add in the current graph the CDF of a Weibull law defined by theses parameters ${\cal W}( lambda, k)$ .

where the CDF formula is : $F(x) = 1 - \exp\left(-(x/\lambda)^k\right) \forall x \ge 0$

Parameters

lambda	(Double_t) The scale parameter $\beta$ of the Weibull law
k	(Double_t) The shape parameter of the the Weibull law
color	(Color_t)[kBlack] The color of the line of the CDF's law
stitle	(TString)[] The title of the entry in the legend box

Referenced by ClassImp().

◆ changeLog()

void URANIE::UncertModeler::TTester::changeLog ( )

inline

References _blog.

◆ computeModifiedScore()

virtual void URANIE::UncertModeler::TTester::computeModifiedScore ( URANIE::DataServer::TStochasticAttribute::ELawType tlaw )

pure virtual

Compute the modified value of the statistic.

Implemented in URANIE::UncertModeler::TTestAndersonDarling, URANIE::UncertModeler::TTestCramerVonMises, and URANIE::UncertModeler::TTestKolmogorovSmirnov.

Referenced by ClassImp().

◆ computePValue()

void URANIE::UncertModeler::TTester::computePValue ( URANIE::DataServer::TStochasticAttribute::ELawType tlaw )

Compute the PValue of the test.

Parameters

tlaw	Type of Law to test

Referenced by ClassImp().

◆ computePValue_AD_Case0()

Double_t URANIE::UncertModeler::TTester::computePValue_AD_Case0 ( Double_t dz )

Compute the PValue of the test AD in case Case0 (all parameters of the PDF are given)

Internal function. These function comes from the Table 4.3 Page 108 of the "Goodness-of-Fit Techniques" book Distribution of $A^{2}$ , Case 0 : The table gives $q = I\!\!P [ A^{2} > z ]$ . Here, we return the Pvalue = .

For the AD test, the AD value is not modified (

See also: Table 4.2 page 105 : $A^{2} \forall n \ge 5$ )

Parameters

dz	The $A^{2}$ value of the AD test (not modified)

Returns: the PValue

Referenced by ClassImp().

◆ computePValue_AD_Case3()

Double_t URANIE::UncertModeler::TTester::computePValue_AD_Case3	(	Double_t	dz,
		URANIE::DataServer::TStochasticAttribute::ELawType	tlaw
	)

Compute the PValue of the test AD in case Case3 (all parameters of the PDF are estimated by MLE)

Parameters

dz	The $A^{2}$ value of the AD test (not modified)
tlaw	The type of Law

Returns: the PValue

Referenced by ClassImp().

◆ computePValue_CvM_Case3()

Double_t URANIE::UncertModeler::TTester::computePValue_CvM_Case3	(	Double_t	dz,
		URANIE::DataServer::TStochasticAttribute::ELawType	tlaw
	)

Compute the PValue of the test CvM in case Case3 (all parameters of the PDF are estimated by MLE)

Parameters

dz	The $W^{2}$ value of the CvM test (not modified)
tlaw	The type of Law

Returns: the PValue

Referenced by ClassImp().

◆ computePValue_KS_CaseAll()

Double_t URANIE::UncertModeler::TTester::computePValue_KS_CaseAll ( Double_t dz )

Compute the PValue of the test KS in all cases (Case0 && Case3)

Parameters

dz	The $D^{\star}$ value of the KS test (modified score)

Returns: The PValue

Referenced by ClassImp().

◆ computeScore()

Double_t URANIE::UncertModeler::TTester::computeScore ( Option_t * option = "" )

Compute the score of the statistic.

A test to generalize the compute the score of the statistic.

Referenced by ClassImp().

◆ computeTheCurrentScore()

virtual Double_t URANIE::UncertModeler::TTester::computeTheCurrentScore	(	Int_t	i,
		Double_t	dzi
	)

pure virtual

The true computation with a vector of information and return the current value.

Parameters

i	(Int_t) the current index
dzi	(Double_t) the current value of Z = F(x)

Implemented in URANIE::UncertModeler::TTestAndersonDarling, URANIE::UncertModeler::TTestCramerVonMises, and URANIE::UncertModeler::TTestKolmogorovSmirnov.

Referenced by ClassImp().

◆ createGraph()

TGraph * URANIE::UncertModeler::TTester::createGraph ( Color_t color = kBlack )

protected

Referenced by ClassImp().

◆ findParametersLaw()

void URANIE::UncertModeler::TTester::findParametersLaw	(	TString	soption,
		TVectorD &	vec
	)

protected

Finds the parameters in the option string.

Example of string parameters are

"normal(-125.025, 125)"
"normal(125.25, .235)"

Referenced by ClassImp().

◆ getCase()

ECase URANIE::UncertModeler::TTester::getCase ( )

inline

getCaseOfGoF

kCase0 : All the parameters of the law are specified kCase3 : All the parameters of the law are not specified. Then they are estimated by MLE.

The other cases are not treated in URANIE: kCase1 : Just the position parameters are specified, but not the scaled parameters kCase2 : Just the scaled parameters are specified, but not the position parameters

Returns: ECase the case of Goodness-Of-Fit

References _nCase.

◆ getID()

Int_t URANIE::UncertModeler::TTester::getID ( )

inline

Referenced by ClassImp().

◆ getLog()

Bool_t URANIE::UncertModeler::TTester::getLog ( )

inline

References _blog.

◆ getModifiedScore()

Double_t URANIE::UncertModeler::TTester::getModifiedScore ( )

inline

Get the modified score of the last test.

References _dModifiedScore.

◆ getPValue()

Double_t URANIE::UncertModeler::TTester::getPValue ( )

inline

Get the pValue of the last test.

References _pValue.

◆ getScore()

Double_t URANIE::UncertModeler::TTester::getScore ( )

inline

Get the score of the last test.

References _dScore.

◆ getTitleStatistic()

virtual const char * URANIE::UncertModeler::TTester::getTitleStatistic ( )

pure virtual

Return the name of the statistic.

Implemented in URANIE::UncertModeler::TTestAndersonDarling, URANIE::UncertModeler::TTestCramerVonMises, and URANIE::UncertModeler::TTestKolmogorovSmirnov.

Referenced by ClassImp().

◆ nTimesValueLTvalVector()

Int_t URANIE::UncertModeler::TTester::nTimesValueLTvalVector	(	Double_t	dval,
		TVectorD	vecValues
	)

protected

Referenced by ClassImp().

◆ printLog()

virtual void URANIE::UncertModeler::TTester::printLog ( Option_t * option = "" )

virtual

Referenced by ClassImp().

◆ setLog()

void URANIE::UncertModeler::TTester::setLog ( )

inline

References _blog.

◆ treatOption()

void URANIE::UncertModeler::TTester::treatOption ( Option_t * option )

Treats the options of the computeScore method.

The option of the computeScore method can be :

"normal"
"normal:normal(200,35)" to test the normal law with $(\mu, \sigma) = (200, 35)$

Referenced by ClassImp().

◆ unsetLog()

void URANIE::UncertModeler::TTester::unsetLog ( )

inline

References _blog.

Member Data Documentation

◆ _bDrawSame

Bool_t URANIE::UncertModeler::TTester::_bDrawSame

protected

Boolean for drawing in the current pad.

Referenced by ClassImp().

◆ _blog

Bool_t URANIE::UncertModeler::TTester::_blog

Boolean for edit the log.

Referenced by changeLog(), ClassImp(), ClassImp(), ClassImp(), ClassImp(), getLog(), setLog(), and unsetLog().

◆ _bquiet

Bool_t URANIE::UncertModeler::TTester::_bquiet

Boolean for being quiet.

Referenced by ClassImp().

◆ _dModifiedScore

Double_t URANIE::UncertModeler::TTester::_dModifiedScore

protected

The Modified value score of the last test.

Referenced by ClassImp(), ClassImp(), ClassImp(), ClassImp(), and getModifiedScore().

◆ _dScore

Double_t URANIE::UncertModeler::TTester::_dScore

protected

The score of the last test.

Referenced by ClassImp(), ClassImp(), ClassImp(), ClassImp(), and getScore().

◆ _listOfLaws

TList* URANIE::UncertModeler::TTester::_listOfLaws

protected

The list of law to test.

Referenced by ClassImp().

◆ _nCase

ECase URANIE::UncertModeler::TTester::_nCase

Referenced by ClassImp(), ClassImp(), ClassImp(), ClassImp(), and getCase().

◆ _nS

Long64_t URANIE::UncertModeler::TTester::_nS

protected

the size of selected values

Referenced by ClassImp(), ClassImp(), ClassImp(), and ClassImp().

◆ _nTestType

ETestType URANIE::UncertModeler::TTester::_nTestType

The type of test.

Referenced by ClassImp().

◆ _pValue

Double_t URANIE::UncertModeler::TTester::_pValue

protected

The pValue of the last test.

Referenced by ClassImp(), and getPValue().

◆ _sAttributes

TString URANIE::UncertModeler::TTester::_sAttributes

protected

The string of attributes separated by the colon ":" character.

Referenced by ClassImp().

◆ _tds

URANIE::DataServer::TDataServer* URANIE::UncertModeler::TTester::_tds

Pointer to a TDS.

Referenced by ClassImp().

Documentation / Manuel développeur

Modules disponibles

Public Types

Public Member Functions

Public Attributes

Protected Member Functions

Protected Attributes

Detailed Description

Member Enumeration Documentation

◆ ECase

◆ ETestType

Constructor & Destructor Documentation

◆ TTester()

◆ ~TTester()

Member Function Documentation

◆ addBetaLaw()

◆ addGammaLaw()

◆ addLogNormalLaw()

◆ addNormalLaw()

◆ addUniformLaw()

◆ addWeibullLaw()

◆ changeLog()

◆ computeModifiedScore()

◆ computePValue()

◆ computePValue_AD_Case0()

◆ computePValue_AD_Case3()

◆ computePValue_CvM_Case3()

◆ computePValue_KS_CaseAll()

◆ computeScore()

◆ computeTheCurrentScore()

◆ createGraph()

◆ findParametersLaw()

◆ getCase()

◆ getID()

◆ getLog()

◆ getModifiedScore()

◆ getPValue()

◆ getScore()

◆ getTitleStatistic()

◆ nTimesValueLTvalVector()

◆ printLog()

◆ setLog()

◆ treatOption()

◆ unsetLog()

Member Data Documentation

◆ _bDrawSame

◆ _blog

◆ _bquiet

◆ _dModifiedScore

◆ _dScore

◆ _listOfLaws

◆ _nCase

◆ _nS

◆ _nTestType

◆ _pValue

◆ _sAttributes

◆ _tds