ToyMCImportanceSampler is an extension of the ToyMCSampler for Importance Sampling.

Implementation based on a work by Cranmer, Kreiss, Read (in Preparation)

Definition at line 22 of file ToyMCImportanceSampler.h.

Public Member Functions
	ToyMCImportanceSampler (TestStatistic &ts, Int_t ntoys)
	~ToyMCImportanceSampler () override
void	AddImportanceDensity (RooAbsPdf p, const RooArgSet s)
	For importance sampling with multiple densities/snapshots: This is used to check the current Likelihood against Likelihoods from other importance densities apart from the one given as importance snapshot.
void	AddNullDensity (RooAbsPdf p, const RooArgSet s=nullptr)
	The pdf can be nullptr in which case the density from SetPdf() is used.
virtual void	AddTestStatistic (TestStatistic *t=nullptr)
	The pdf can be nullptr in which case the density from SetPdf() is used.
virtual SamplingDistribution *	AppendSamplingDistribution (RooArgSet &allParameters, SamplingDistribution *last, Int_t additionalMC)
	Extended interface to append to sampling distribution more samples.
bool	CheckConfig (void)
	Checks for sufficient information to do a GetSamplingDistribution(...).
double	ConfidenceLevel () const override
	Get the Confidence level for the test.
int	CreateImpDensitiesForOnePOIAdaptively (RooAbsPdf &pdf, const RooArgSet &allPOI, RooRealVar &poi, double nStdDevOverlap=0.5, double poiValueForBackground=0.0)
	poi has to be fitted beforehand. This function expects this to be the muhat value.
int	CreateNImpDensitiesForOnePOI (RooAbsPdf &pdf, const RooArgSet &allPOI, RooRealVar &poi, int n, double poiValueForBackground=0.0)
	n is the number of importance densities
virtual RooArgList *	EvaluateAllTestStatistics (RooAbsData &data, const RooArgSet &poi)
	Evaluate all test statistics, returning result and any detailed output.
double	EvaluateTestStatistic (RooAbsData &data, RooArgSet &nullPOI) override
	Main interface to evaluate the test statistic on a dataset.
virtual double	EvaluateTestStatistic (RooAbsData &data, RooArgSet &nullPOI, int i)
	Main interface to evaluate the test statistic on a dataset.
virtual void	GenerateGlobalObservables (RooAbsPdf &pdf) const
	generate global observables
virtual RooAbsData *	GenerateToyData (RooArgSet &paramPoint) const
RooAbsData *	GenerateToyData (RooArgSet &paramPoint, double &weight) const override
virtual RooAbsData *	GenerateToyData (RooArgSet &paramPoint, double &weight, RooAbsPdf &pdf) const
	generates toy data with weight
virtual RooAbsData *	GenerateToyData (RooArgSet &paramPoint, double &weight, std::vector< double > &impNLLs, double &nullNLL) const
virtual RooAbsData *	GenerateToyData (RooArgSet &paramPoint, RooAbsPdf &pdf) const
	generates toy data without weight
virtual RooAbsData *	GenerateToyData (std::vector< double > &weights) const
virtual RooAbsData *	GenerateToyData (std::vector< double > &weights, std::vector< double > &nullNLLs, std::vector< double > &impNLLs) const
	This method generates a toy data set for importance sampling for the given parameter point taking global observables into account.
virtual Int_t	GetNToys (void)
std::string	GetSamplingDistName (void)
SamplingDistribution *	GetSamplingDistribution (RooArgSet &paramPoint) override
	main interface
virtual RooDataSet *	GetSamplingDistributions (RooArgSet &paramPoint)
	Use for serial and parallel runs.
RooDataSet *	GetSamplingDistributionsSingleWorker (RooArgSet &paramPoint) override
	overwrite GetSamplingDistributionsSingleWorker(paramPoint) with a version that loops over nulls and importance densities, but calls the parent ToyMCSampler::GetSamplingDistributionsSingleWorker(paramPoint).
virtual TestStatistic *	GetTestStatistic (unsigned int i) const
TestStatistic *	GetTestStatistic (void) const override
	Get the TestStatistic.
void	Initialize (RooAbsArg &, RooArgSet &, RooArgSet &) override
	Common Initialization.
TClass *	IsA () const override
int	nEventsPerToy () const
void	SetApplyVeto (bool b=true)
	When set to true, this sets the weight of all toys to zero that do not have the largest likelihood under the density it was generated compared to the other densities.
virtual void	SetAsimovNuisancePar (bool i=true)
virtual void	SetConditionalObservables (const RooArgSet &set)
	set the conditional observables which will be used when creating the NLL so the pdf's will not be normalized on the conditional observables when computing the NLL Since the class use a NLL we need to set the conditional observables if they exist in the model
void	SetConfidenceLevel (double cl) override
	set the confidence level for the interval (eg. 0.95 for a 95% Confidence Interval)
void	SetDensityToGenerateFromByIndex (unsigned int i, bool fromNull=false)
	specifies the pdf to sample from
void	SetEqualNumToysPerDensity (void)
virtual void	SetExpectedNuisancePar (bool i=true)
void	SetExpIncreasingNumToysPerDensity (void)
void	SetGenerateAutoBinned (bool autoBinned=true)
	set auto binned generation (=> see RooFit::AutoBinned() option)
void	SetGenerateBinned (bool binned=true)
	control to use bin data generation (=> see RooFit::AllBinned() option)
void	SetGenerateBinnedTag (const char *binnedTag="")
	name of the tag for individual components to be generated binned (=> see RooFit::GenBinned() option)
void	SetGlobalObservables (const RooArgSet &o) override
	specify the conditional observables
void	SetMaxToys (double t)
	This option forces a maximum number of total toys.
virtual void	SetNEventsPerToy (const Int_t nevents)
	Forces the generation of exactly n events even for extended PDFs.
virtual void	SetNToys (const Int_t ntoy)
void	SetNuisanceParameters (const RooArgSet &np) override
	specify the nuisance parameters (eg. the rest of the parameters)
void	SetObservables (const RooArgSet &o) override
	specify the observables in the dataset (needed to evaluate the test statistic)
void	SetParametersForTestStat (const RooArgSet &nullpoi) override
	overwrite from ToyMCSampler
void	SetPdf (RooAbsPdf &pdf) override
	overwrite from ToyMCSampler
void	SetPriorNuisance (RooAbsPdf *pdf) override
	How to randomize the prior. Set to nullptr to deactivate randomization.
void	SetProtoData (const RooAbsData *d)
void	SetReuseNLL (bool r=true)
void	SetSamplingDistName (const char *name) override
	Set the name of the sampling distribution used for plotting.
void	SetTestSize (double size) override
	set the size of the test (rate of Type I error) ( Eg. 0.05 for a 95% Confidence Interval)
void	SetTestStatistic (TestStatistic *t) override
	Set the TestStatistic (want the argument to be a function of the data & parameter points.
virtual void	SetTestStatistic (TestStatistic *testStatistic, unsigned int i)
	Set the TestStatistic (want the argument to be a function of the data & parameter points.
void	SetToysBothTails (double toys, double low_threshold, double high_threshold)
void	SetToysLeftTail (double toys, double threshold)
void	SetToysRightTail (double toys, double threshold)
void	SetUseMultiGen (bool flag)
void	Streamer (TBuffer &) override
void	StreamerNVirtual (TBuffer &ClassDef_StreamerNVirtual_b)

Static Public Member Functions
static TClass *	Class ()
static const char *	Class_Name ()
static constexpr Version_t	Class_Version ()
static const char *	DeclFileName ()
static void	SetAlwaysUseMultiGen (bool flag)

Protected Member Functions
void	ClearCache () override
	helper method for clearing the cache
const RooArgList *	EvaluateAllTestStatistics (RooAbsData &data, const RooArgSet &poi, DetailedOutputAggregator &detOutAgg)
std::unique_ptr< RooAbsData >	Generate (RooAbsPdf &pdf, RooArgSet &observables, const RooAbsData *protoData=nullptr, int forceEvents=0) const
	helper for GenerateToyData

Protected Attributes
std::unique_ptr< RooArgSet >	_allVars
	!
std::unique_ptr< RooAbsPdf::GenSpec >	_gs1
	! GenSpec #1
std::unique_ptr< RooAbsPdf::GenSpec >	_gs2
	! GenSpec #2
std::unique_ptr< RooAbsPdf::GenSpec >	_gs3
	! GenSpec #3
std::unique_ptr< RooAbsPdf::GenSpec >	_gs4
	! GenSpec #4
std::vector< std::unique_ptr< RooAbsPdf::GenSpec > >	_gsList
	!
std::vector< std::unique_ptr< RooArgSet > >	_obsList
	!
std::vector< RooAbsPdf * >	_pdfList
	! We don't own those objects
double	fAdaptiveHighLimit
double	fAdaptiveLowLimit
	tails
bool	fApplyVeto = true
RooArgSet	fConditionalObs
	set of conditional observables
bool	fExpectedNuisancePar
	whether to use expectation values for nuisance parameters (ie Asimov data set)
bool	fGenerateAutoBinned = true
bool	fGenerateBinned = false
TString	fGenerateBinnedTag = ""
bool	fGenerateFromNull = true
const RooArgSet *	fGlobalObservables = nullptr
std::vector< std::unique_ptr< RooAbsReal > >	fImpNLLs
	!
std::vector< RooAbsPdf * >	fImportanceDensities
std::vector< const RooArgSet * >	fImportanceSnapshots
unsigned int	fIndexGenDensity = 0
double	fMaxToys
	maximum no of toys (taking weights into account, therefore double)
Int_t	fNEvents = 0
	number of events per toy (may be ignored depending on settings)
Int_t	fNToys
	number of toys to generate
NuisanceParametersSampler *	fNuisanceParametersSampler = nullptr
	!
const RooArgSet *	fNuisancePars = nullptr
std::vector< RooAbsPdf * >	fNullDensities
	support multiple null densities
std::vector< std::unique_ptr< RooAbsReal > >	fNullNLLs
	!
std::vector< const RooArgSet * >	fNullSnapshots
const RooArgSet *	fObservables = nullptr
std::unique_ptr< const RooArgSet >	fParametersForTestStat
RooAbsPdf *	fPdf = nullptr
	densities, snapshots, and test statistics to reweight to
RooAbsPdf *	fPriorNuisance = nullptr
	prior pdf for nuisance parameters
const RooAbsData *	fProtoData = nullptr
	in dev
bool	fReuseNLL = true
std::string	fSamplingDistName
	name of the model
double	fSize = 0.05
std::vector< TestStatistic * >	fTestStatistics
double	fToysInTails = 0.0
	minimum no of toys in tails for adaptive sampling (taking weights into account, therefore double) Default: 0.0 which means no adaptive sampling
toysStrategies	fToysStrategy = EQUALTOYSPERDENSITY
bool	fUseMultiGen = false
	Use PrepareMultiGen?

Static Protected Attributes
static bool	fgAlwaysUseMultiGen = false
	Use PrepareMultiGen always.

#include <RooStats/ToyMCImportanceSampler.h>

Inheritance diagram for RooStats::ToyMCImportanceSampler:

Constructor & Destructor Documentation

◆ ToyMCImportanceSampler()

RooStats::ToyMCImportanceSampler::ToyMCImportanceSampler	(	TestStatistic &	ts,
		Int_t	ntoys )

inline

Definition at line 25 of file ToyMCImportanceSampler.h.

◆ ~ToyMCImportanceSampler()

RooStats::ToyMCImportanceSampler::~ToyMCImportanceSampler ( )

override

Definition at line 36 of file ToyMCImportanceSampler.cxx.

Member Function Documentation

◆ AddImportanceDensity()

void RooStats::ToyMCImportanceSampler::AddImportanceDensity	(	RooAbsPdf *	p,
		const RooArgSet *	s )

inline

For importance sampling with multiple densities/snapshots: This is used to check the current Likelihood against Likelihoods from other importance densities apart from the one given as importance snapshot.

The pdf can be nullptr in which case the density from SetImportanceDensity() is used. The snapshot is also optional.

Definition at line 62 of file ToyMCImportanceSampler.h.

◆ AddNullDensity()

void RooStats::ToyMCImportanceSampler::AddNullDensity	(	RooAbsPdf *	p,
		const RooArgSet *	s = nullptr )

inline

The pdf can be nullptr in which case the density from SetPdf() is used.

The snapshot and TestStatistic is also optional.

Definition at line 83 of file ToyMCImportanceSampler.h.

◆ AddTestStatistic()

virtual void RooStats::ToyMCSampler::AddTestStatistic ( TestStatistic * t = nullptr )

inlinevirtualinherited

The pdf can be nullptr in which case the density from SetPdf() is used.

The snapshot and TestStatistic is also optional.

Definition at line 91 of file ToyMCSampler.h.

◆ AppendSamplingDistribution()

SamplingDistribution * RooStats::ToyMCSampler::AppendSamplingDistribution	(	RooArgSet &	allParameters,
		SamplingDistribution *	last,
		Int_t	additionalMC )

virtualinherited

Extended interface to append to sampling distribution more samples.

Definition at line 551 of file ToyMCSampler.cxx.

◆ CheckConfig()

bool RooStats::ToyMCSampler::CheckConfig ( void )

inherited

Checks for sufficient information to do a GetSamplingDistribution(...).

only checks, no guessing/determination (do this in calculators, e.g.

using ModelConfig::GuessObsAndNuisance(...))

Definition at line 168 of file ToyMCSampler.cxx.

◆ Class()

TClass * RooStats::ToyMCImportanceSampler::Class ( )

static

Returns: TClass describing this class

◆ Class_Name()

const char * RooStats::ToyMCImportanceSampler::Class_Name ( )

static

Returns: Name of this class

◆ Class_Version()

constexpr Version_t RooStats::ToyMCImportanceSampler::Class_Version ( )

inlinestaticconstexpr

Returns: Version of this class

Definition at line 177 of file ToyMCImportanceSampler.h.

◆ ClearCache()

void RooStats::ToyMCImportanceSampler::ClearCache ( void )

overrideprotectedvirtual

helper method for clearing the cache

Reimplemented from RooStats::ToyMCSampler.

Definition at line 43 of file ToyMCImportanceSampler.cxx.

◆ ConfidenceLevel()

double RooStats::ToyMCSampler::ConfidenceLevel ( ) const

inlineoverridevirtualinherited

Get the Confidence level for the test.

Implements RooStats::TestStatSampler.

Definition at line 131 of file ToyMCSampler.h.

◆ CreateImpDensitiesForOnePOIAdaptively()

int RooStats::ToyMCImportanceSampler::CreateImpDensitiesForOnePOIAdaptively	(	RooAbsPdf &	pdf,
		const RooArgSet &	allPOI,
		RooRealVar &	poi,
		double	nStdDevOverlap = 0.5,
		double	poiValueForBackground = 0.0 )

poi has to be fitted beforehand. This function expects this to be the muhat value.

Definition at line 427 of file ToyMCImportanceSampler.cxx.

◆ CreateNImpDensitiesForOnePOI()

int RooStats::ToyMCImportanceSampler::CreateNImpDensitiesForOnePOI	(	RooAbsPdf &	pdf,
		const RooArgSet &	allPOI,
		RooRealVar &	poi,
		int	n,
		double	poiValueForBackground = 0.0 )

n is the number of importance densities

Definition at line 450 of file ToyMCImportanceSampler.cxx.

◆ DeclFileName()

const char * RooStats::ToyMCImportanceSampler::DeclFileName ( )

inlinestatic

Returns: Name of the file containing the class declaration

Definition at line 177 of file ToyMCImportanceSampler.h.

◆ EvaluateAllTestStatistics() [1/2]

RooArgList * RooStats::ToyMCSampler::EvaluateAllTestStatistics	(	RooAbsData &	data,
		const RooArgSet &	poi )

virtualinherited

Evaluate all test statistics, returning result and any detailed output.

PDF parameter values are saved in case they are modified by TestStatistic::Evaluate (eg. SimpleLikelihoodRatioTestStat).

Definition at line 184 of file ToyMCSampler.cxx.

◆ EvaluateAllTestStatistics() [2/2]

const RooArgList * RooStats::ToyMCSampler::EvaluateAllTestStatistics	(	RooAbsData &	data,
		const RooArgSet &	poi,
		DetailedOutputAggregator &	detOutAgg )

protectedinherited

Definition at line 194 of file ToyMCSampler.cxx.

◆ EvaluateTestStatistic() [1/2]

double RooStats::ToyMCSampler::EvaluateTestStatistic	(	RooAbsData &	data,
		RooArgSet &	paramsOfInterest )

inlineoverridevirtualinherited

Main interface to evaluate the test statistic on a dataset.

Implements RooStats::TestStatSampler.

Definition at line 121 of file ToyMCSampler.h.

◆ EvaluateTestStatistic() [2/2]

virtual double RooStats::ToyMCSampler::EvaluateTestStatistic	(	RooAbsData &	data,
		RooArgSet &	nullPOI,
		int	i )

inlinevirtualinherited

Main interface to evaluate the test statistic on a dataset.

Definition at line 118 of file ToyMCSampler.h.

◆ Generate()

std::unique_ptr< RooAbsData > RooStats::ToyMCSampler::Generate	(	RooAbsPdf &	pdf,
		RooArgSet &	observables,
		const RooAbsData *	protoData = nullptr,
		int	forceEvents = 0 ) const

protectedinherited

helper for GenerateToyData

This is the generate function to use in the context of the ToyMCSampler instead of the standard RooAbsPdf::generate(...).

It takes into account whether the number of events is given explicitly or whether it should use the expected number of events. It also takes into account the option to generate a binned data set (i.e. RooDataHist).

Definition at line 479 of file ToyMCSampler.cxx.

◆ GenerateGlobalObservables()

void RooStats::ToyMCSampler::GenerateGlobalObservables ( RooAbsPdf & pdf ) const

virtualinherited

generate global observables

Definition at line 347 of file ToyMCSampler.cxx.

◆ GenerateToyData() [1/7]

virtual RooAbsData * RooStats::ToyMCSampler::GenerateToyData ( RooArgSet & paramPoint ) const

inlinevirtual

Reimplemented from RooStats::ToyMCSampler.

Definition at line 107 of file ToyMCSampler.h.

◆ GenerateToyData() [2/7]

RooAbsData * RooStats::ToyMCImportanceSampler::GenerateToyData	(	RooArgSet &	paramPoint,
		double &	weight ) const

overridevirtual

Reimplemented from RooStats::ToyMCSampler.

Definition at line 128 of file ToyMCImportanceSampler.cxx.

◆ GenerateToyData() [3/7]

RooAbsData * RooStats::ToyMCSampler::GenerateToyData	(	RooArgSet &	paramPoint,
		double &	weight,
		RooAbsPdf &	pdf ) const

virtual

generates toy data with weight

This method generates a toy data set for the given parameter point taking global observables into account.

The values of the generated global observables remain in the pdf's variables. They have to have those values for the subsequent evaluation of the test statistics.

Reimplemented from RooStats::ToyMCSampler.

Definition at line 110 of file ToyMCSampler.cxx.

◆ GenerateToyData() [4/7]

RooAbsData * RooStats::ToyMCImportanceSampler::GenerateToyData	(	RooArgSet &	paramPoint,
		double &	weight,
		std::vector< double > &	impNLLs,
		double &	nullNLL ) const

virtual

Definition at line 170 of file ToyMCImportanceSampler.cxx.

◆ GenerateToyData() [5/7]

virtual RooAbsData * RooStats::ToyMCSampler::GenerateToyData	(	RooArgSet &	paramPoint,
		RooAbsPdf &	pdf ) const

inlinevirtual

generates toy data without weight

Reimplemented from RooStats::ToyMCSampler.

Definition at line 102 of file ToyMCSampler.h.

◆ GenerateToyData() [6/7]

RooAbsData * RooStats::ToyMCImportanceSampler::GenerateToyData ( std::vector< double > & weights ) const

virtual

Definition at line 209 of file ToyMCImportanceSampler.cxx.

◆ GenerateToyData() [7/7]

RooAbsData * RooStats::ToyMCImportanceSampler::GenerateToyData	(	std::vector< double > &	weights,
		std::vector< double > &	nullNLLs,
		std::vector< double > &	impNLLs ) const

virtual

This method generates a toy data set for importance sampling for the given parameter point taking global observables into account.

The values of the generated global observables remain in the pdf's variables. They have to have those values for the subsequent evaluation of the test statistics.

Definition at line 233 of file ToyMCImportanceSampler.cxx.

◆ GetNToys()

virtual Int_t RooStats::ToyMCSampler::GetNToys ( void )

inlinevirtualinherited

Definition at line 138 of file ToyMCSampler.h.

◆ GetSamplingDistName()

std::string RooStats::ToyMCSampler::GetSamplingDistName ( void )

inlineinherited

Definition at line 211 of file ToyMCSampler.h.

◆ GetSamplingDistribution()

SamplingDistribution * RooStats::ToyMCSampler::GetSamplingDistribution ( RooArgSet & paramPoint )

overridevirtualinherited

main interface

Implements RooStats::TestStatSampler.

Definition at line 226 of file ToyMCSampler.cxx.

◆ GetSamplingDistributions()

RooDataSet * RooStats::ToyMCSampler::GetSamplingDistributions ( RooArgSet & paramPoint )

virtualinherited

Use for serial and parallel runs.

Definition at line 246 of file ToyMCSampler.cxx.

◆ GetSamplingDistributionsSingleWorker()

RooDataSet * RooStats::ToyMCImportanceSampler::GetSamplingDistributionsSingleWorker ( RooArgSet & paramPoint )

overridevirtual

overwrite GetSamplingDistributionsSingleWorker(paramPoint) with a version that loops over nulls and importance densities, but calls the parent ToyMCSampler::GetSamplingDistributionsSingleWorker(paramPoint).

Reimplemented from RooStats::ToyMCSampler.

Definition at line 52 of file ToyMCImportanceSampler.cxx.

◆ GetTestStatistic() [1/2]

virtual TestStatistic * RooStats::ToyMCSampler::GetTestStatistic ( unsigned int i ) const

inlinevirtualinherited

Definition at line 125 of file ToyMCSampler.h.

◆ GetTestStatistic() [2/2]

TestStatistic * RooStats::ToyMCSampler::GetTestStatistic ( void ) const

inlineoverridevirtualinherited

Get the TestStatistic.

Implements RooStats::TestStatSampler.

Definition at line 129 of file ToyMCSampler.h.

◆ Initialize()

void RooStats::ToyMCSampler::Initialize	(	RooAbsArg &	testStatistic,
		RooArgSet &	paramsOfInterest,
		RooArgSet &	nuisanceParameters )

inlineoverridevirtualinherited

Common Initialization.

Implements RooStats::TestStatSampler.

Definition at line 132 of file ToyMCSampler.h.

◆ IsA()

TClass * RooStats::ToyMCImportanceSampler::IsA ( ) const

inlineoverridevirtual

Returns: TClass describing current object

Reimplemented from RooStats::TestStatSampler.

Definition at line 177 of file ToyMCImportanceSampler.h.

◆ nEventsPerToy()

int RooStats::ToyMCSampler::nEventsPerToy ( ) const

inlineinherited

Definition at line 147 of file ToyMCSampler.h.

◆ SetAlwaysUseMultiGen()

void RooStats::ToyMCSampler::SetAlwaysUseMultiGen ( bool flag )

staticinherited

Definition at line 138 of file ToyMCSampler.cxx.

◆ SetApplyVeto()

void RooStats::ToyMCImportanceSampler::SetApplyVeto ( bool b = true )

inline

When set to true, this sets the weight of all toys to zero that do not have the largest likelihood under the density it was generated compared to the other densities.

Definition at line 124 of file ToyMCImportanceSampler.h.

◆ SetAsimovNuisancePar()

virtual void RooStats::ToyMCSampler::SetAsimovNuisancePar ( bool i = true )

inlinevirtualinherited

Definition at line 197 of file ToyMCSampler.h.

◆ SetConditionalObservables()

virtual void RooStats::ToyMCImportanceSampler::SetConditionalObservables ( const RooArgSet & set )

inlinevirtual

set the conditional observables which will be used when creating the NLL so the pdf's will not be normalized on the conditional observables when computing the NLL Since the class use a NLL we need to set the conditional observables if they exist in the model

Definition at line 131 of file ToyMCImportanceSampler.h.

◆ SetConfidenceLevel()

void RooStats::ToyMCSampler::SetConfidenceLevel ( double cl )

inlineoverridevirtualinherited

set the confidence level for the interval (eg. 0.95 for a 95% Confidence Interval)

Implements RooStats::TestStatSampler.

Definition at line 180 of file ToyMCSampler.h.

◆ SetDensityToGenerateFromByIndex()

void RooStats::ToyMCImportanceSampler::SetDensityToGenerateFromByIndex	(	unsigned int	i,
		bool	fromNull = false )

inline

specifies the pdf to sample from

Definition at line 42 of file ToyMCImportanceSampler.h.

◆ SetEqualNumToysPerDensity()

void RooStats::ToyMCImportanceSampler::SetEqualNumToysPerDensity ( void )

inline

Definition at line 148 of file ToyMCImportanceSampler.h.

◆ SetExpectedNuisancePar()

virtual void RooStats::ToyMCSampler::SetExpectedNuisancePar ( bool i = true )

inlinevirtualinherited

Definition at line 196 of file ToyMCSampler.h.

◆ SetExpIncreasingNumToysPerDensity()

void RooStats::ToyMCImportanceSampler::SetExpIncreasingNumToysPerDensity ( void )

inline

Definition at line 149 of file ToyMCImportanceSampler.h.

◆ SetGenerateAutoBinned()

void RooStats::ToyMCSampler::SetGenerateAutoBinned ( bool autoBinned = true )

inlineinherited

set auto binned generation (=> see RooFit::AutoBinned() option)

Definition at line 207 of file ToyMCSampler.h.

◆ SetGenerateBinned()

void RooStats::ToyMCSampler::SetGenerateBinned ( bool binned = true )

inlineinherited

control to use bin data generation (=> see RooFit::AllBinned() option)

Definition at line 203 of file ToyMCSampler.h.

◆ SetGenerateBinnedTag()

void RooStats::ToyMCSampler::SetGenerateBinnedTag ( const char * binnedTag = "" )

inlineinherited

name of the tag for individual components to be generated binned (=> see RooFit::GenBinned() option)

Definition at line 205 of file ToyMCSampler.h.

◆ SetGlobalObservables()

void RooStats::ToyMCSampler::SetGlobalObservables ( const RooArgSet & o )

inlineoverridevirtualinherited

specify the conditional observables

Implements RooStats::TestStatSampler.

Definition at line 174 of file ToyMCSampler.h.

◆ SetMaxToys()

void RooStats::ToyMCSampler::SetMaxToys ( double t )

inlineinherited

This option forces a maximum number of total toys.

Definition at line 214 of file ToyMCSampler.h.

◆ SetNEventsPerToy()

virtual void RooStats::ToyMCSampler::SetNEventsPerToy ( const Int_t nevents )

inlinevirtualinherited

Forces the generation of exactly n events even for extended PDFs.

Set to 0 to use the Poisson-distributed events from the extended PDF.

Definition at line 142 of file ToyMCSampler.h.

◆ SetNToys()

virtual void RooStats::ToyMCSampler::SetNToys ( const Int_t ntoy )

inlinevirtualinherited

Definition at line 139 of file ToyMCSampler.h.

◆ SetNuisanceParameters()

void RooStats::ToyMCSampler::SetNuisanceParameters ( const RooArgSet & np )

inlineoverridevirtualinherited

specify the nuisance parameters (eg. the rest of the parameters)

Implements RooStats::TestStatSampler.

Definition at line 170 of file ToyMCSampler.h.

◆ SetObservables()

void RooStats::ToyMCSampler::SetObservables ( const RooArgSet & o )

inlineoverridevirtualinherited

specify the observables in the dataset (needed to evaluate the test statistic)

Implements RooStats::TestStatSampler.

Definition at line 172 of file ToyMCSampler.h.

◆ SetParametersForTestStat()

void RooStats::ToyMCImportanceSampler::SetParametersForTestStat ( const RooArgSet & nullpoi )

inlineoverridevirtual

overwrite from ToyMCSampler

Implements RooStats::TestStatSampler.

Definition at line 109 of file ToyMCImportanceSampler.h.

◆ SetPdf()

void RooStats::ToyMCImportanceSampler::SetPdf ( RooAbsPdf & pdf )

inlineoverridevirtual

overwrite from ToyMCSampler

Implements RooStats::TestStatSampler.

Definition at line 99 of file ToyMCImportanceSampler.h.

◆ SetPriorNuisance()

void RooStats::ToyMCSampler::SetPriorNuisance ( RooAbsPdf * pdf )

inlineoverridevirtualinherited

How to randomize the prior. Set to nullptr to deactivate randomization.

Implements RooStats::TestStatSampler.

Definition at line 162 of file ToyMCSampler.h.

◆ SetProtoData()

void RooStats::ToyMCSampler::SetProtoData ( const RooAbsData * d )

inlineinherited

Definition at line 232 of file ToyMCSampler.h.

◆ SetReuseNLL()

void RooStats::ToyMCImportanceSampler::SetReuseNLL ( bool r = true )

inline

Definition at line 126 of file ToyMCImportanceSampler.h.

◆ SetSamplingDistName()

void RooStats::ToyMCSampler::SetSamplingDistName ( const char * name )

inlineoverridevirtualinherited

Set the name of the sampling distribution used for plotting.

Implements RooStats::TestStatSampler.

Definition at line 210 of file ToyMCSampler.h.

◆ SetTestSize()

void RooStats::ToyMCSampler::SetTestSize ( double size )

inlineoverridevirtualinherited

set the size of the test (rate of Type I error) ( Eg. 0.05 for a 95% Confidence Interval)

Implements RooStats::TestStatSampler.

Definition at line 178 of file ToyMCSampler.h.

◆ SetTestStatistic() [1/2]

void RooStats::ToyMCSampler::SetTestStatistic ( TestStatistic * testStatistic )

inlineoverridevirtualinherited

Set the TestStatistic (want the argument to be a function of the data & parameter points.

Implements RooStats::TestStatSampler.

Definition at line 194 of file ToyMCSampler.h.

◆ SetTestStatistic() [2/2]

virtual void RooStats::ToyMCSampler::SetTestStatistic	(	TestStatistic *	testStatistic,
		unsigned int	i )

inlinevirtualinherited

Set the TestStatistic (want the argument to be a function of the data & parameter points.

Definition at line 183 of file ToyMCSampler.h.

◆ SetToysBothTails()

void RooStats::ToyMCSampler::SetToysBothTails	(	double	toys,
		double	low_threshold,
		double	high_threshold )

inlineinherited

Definition at line 226 of file ToyMCSampler.h.

◆ SetToysLeftTail()

void RooStats::ToyMCSampler::SetToysLeftTail	(	double	toys,
		double	threshold )

inlineinherited

Definition at line 216 of file ToyMCSampler.h.

◆ SetToysRightTail()

void RooStats::ToyMCSampler::SetToysRightTail	(	double	toys,
		double	threshold )

inlineinherited

Definition at line 221 of file ToyMCSampler.h.

◆ SetUseMultiGen()

void RooStats::ToyMCSampler::SetUseMultiGen ( bool flag )

inlineinherited

Definition at line 75 of file ToyMCSampler.h.

◆ Streamer()

void RooStats::ToyMCImportanceSampler::Streamer ( TBuffer & )

overridevirtual

Reimplemented from RooStats::TestStatSampler.

◆ StreamerNVirtual()

void RooStats::ToyMCImportanceSampler::StreamerNVirtual ( TBuffer & ClassDef_StreamerNVirtual_b )

inline

Definition at line 177 of file ToyMCImportanceSampler.h.

Member Data Documentation

◆ _allVars

std::unique_ptr<RooArgSet> RooStats::ToyMCSampler::_allVars

mutableprotectedinherited

!

Definition at line 280 of file ToyMCSampler.h.

◆ _gs1

std::unique_ptr<RooAbsPdf::GenSpec> RooStats::ToyMCSampler::_gs1

mutableprotectedinherited

! GenSpec #1

Definition at line 284 of file ToyMCSampler.h.

◆ _gs2

std::unique_ptr<RooAbsPdf::GenSpec> RooStats::ToyMCSampler::_gs2

mutableprotectedinherited

! GenSpec #2

Definition at line 285 of file ToyMCSampler.h.

◆ _gs3

std::unique_ptr<RooAbsPdf::GenSpec> RooStats::ToyMCSampler::_gs3

mutableprotectedinherited

! GenSpec #3

Definition at line 286 of file ToyMCSampler.h.

◆ _gs4

std::unique_ptr<RooAbsPdf::GenSpec> RooStats::ToyMCSampler::_gs4

mutableprotectedinherited

! GenSpec #4

Definition at line 287 of file ToyMCSampler.h.

◆ _gsList

std::vector<std::unique_ptr<RooAbsPdf::GenSpec> > RooStats::ToyMCSampler::_gsList

mutableprotectedinherited

!

Definition at line 283 of file ToyMCSampler.h.

◆ _obsList

std::vector<std::unique_ptr<RooArgSet> > RooStats::ToyMCSampler::_obsList

mutableprotectedinherited

!

Definition at line 282 of file ToyMCSampler.h.

◆ _pdfList

std::vector<RooAbsPdf*> RooStats::ToyMCSampler::_pdfList

mutableprotectedinherited

! We don't own those objects

Definition at line 281 of file ToyMCSampler.h.

◆ fAdaptiveHighLimit

double RooStats::ToyMCSampler::fAdaptiveHighLimit

protectedinherited

Definition at line 273 of file ToyMCSampler.h.

◆ fAdaptiveLowLimit

double RooStats::ToyMCSampler::fAdaptiveLowLimit

protectedinherited

tails

Definition at line 272 of file ToyMCSampler.h.

◆ fApplyVeto

bool RooStats::ToyMCImportanceSampler::fApplyVeto = true

protected

Definition at line 158 of file ToyMCImportanceSampler.h.

◆ fConditionalObs

RooArgSet RooStats::ToyMCImportanceSampler::fConditionalObs

protected

set of conditional observables

Definition at line 160 of file ToyMCImportanceSampler.h.

◆ fExpectedNuisancePar

bool RooStats::ToyMCSampler::fExpectedNuisancePar

protectedinherited

Initial value:

=

false

whether to use expectation values for nuisance parameters (ie Asimov data set)

Definition at line 258 of file ToyMCSampler.h.

◆ fgAlwaysUseMultiGen

bool RooStats::ToyMCSampler::fgAlwaysUseMultiGen = false

staticprotectedinherited

Use PrepareMultiGen always.

Definition at line 289 of file ToyMCSampler.h.

◆ fGenerateAutoBinned

bool RooStats::ToyMCSampler::fGenerateAutoBinned = true

protectedinherited

Definition at line 262 of file ToyMCSampler.h.

◆ fGenerateBinned

bool RooStats::ToyMCSampler::fGenerateBinned = false

protectedinherited

Definition at line 260 of file ToyMCSampler.h.

◆ fGenerateBinnedTag

TString RooStats::ToyMCSampler::fGenerateBinnedTag = ""

protectedinherited

Definition at line 261 of file ToyMCSampler.h.

◆ fGenerateFromNull

bool RooStats::ToyMCImportanceSampler::fGenerateFromNull = true

protected

Definition at line 157 of file ToyMCImportanceSampler.h.

◆ fGlobalObservables

const RooArgSet* RooStats::ToyMCSampler::fGlobalObservables = nullptr

protectedinherited

Definition at line 254 of file ToyMCSampler.h.

◆ fImpNLLs

std::vector<std::unique_ptr<RooAbsReal> > RooStats::ToyMCImportanceSampler::fImpNLLs

mutableprotected

!

Definition at line 175 of file ToyMCImportanceSampler.h.

◆ fImportanceDensities

std::vector<RooAbsPdf*> RooStats::ToyMCImportanceSampler::fImportanceDensities

protected

Definition at line 167 of file ToyMCImportanceSampler.h.

◆ fImportanceSnapshots

std::vector<const RooArgSet*> RooStats::ToyMCImportanceSampler::fImportanceSnapshots

protected

Definition at line 168 of file ToyMCImportanceSampler.h.

◆ fIndexGenDensity

unsigned int RooStats::ToyMCImportanceSampler::fIndexGenDensity = 0

protected

Definition at line 156 of file ToyMCImportanceSampler.h.

◆ fMaxToys

double RooStats::ToyMCSampler::fMaxToys

protectedinherited

maximum no of toys (taking weights into account, therefore double)

Definition at line 270 of file ToyMCSampler.h.

◆ fNEvents

Int_t RooStats::ToyMCSampler::fNEvents = 0

protectedinherited

number of events per toy (may be ignored depending on settings)

Definition at line 256 of file ToyMCSampler.h.

◆ fNToys

Int_t RooStats::ToyMCSampler::fNToys

protectedinherited

number of toys to generate

Definition at line 255 of file ToyMCSampler.h.

◆ fNuisanceParametersSampler

NuisanceParametersSampler* RooStats::ToyMCSampler::fNuisanceParametersSampler = nullptr

mutableprotectedinherited

!

Definition at line 277 of file ToyMCSampler.h.

◆ fNuisancePars

const RooArgSet* RooStats::ToyMCSampler::fNuisancePars = nullptr

protectedinherited

Definition at line 252 of file ToyMCSampler.h.

◆ fNullDensities

std::vector<RooAbsPdf*> RooStats::ToyMCImportanceSampler::fNullDensities

protected

support multiple null densities

Definition at line 163 of file ToyMCImportanceSampler.h.

◆ fNullNLLs

std::vector<std::unique_ptr<RooAbsReal> > RooStats::ToyMCImportanceSampler::fNullNLLs

mutableprotected

!

Definition at line 174 of file ToyMCImportanceSampler.h.

◆ fNullSnapshots

std::vector<const RooArgSet*> RooStats::ToyMCImportanceSampler::fNullSnapshots

mutableprotected

Definition at line 164 of file ToyMCImportanceSampler.h.

◆ fObservables

const RooArgSet* RooStats::ToyMCSampler::fObservables = nullptr

protectedinherited

Definition at line 253 of file ToyMCSampler.h.

◆ fParametersForTestStat

std::unique_ptr<const RooArgSet> RooStats::ToyMCSampler::fParametersForTestStat

protectedinherited

Definition at line 247 of file ToyMCSampler.h.

◆ fPdf

RooAbsPdf* RooStats::ToyMCSampler::fPdf = nullptr

protectedinherited

densities, snapshots, and test statistics to reweight to

model (can be alt or null)

Definition at line 246 of file ToyMCSampler.h.

◆ fPriorNuisance

RooAbsPdf* RooStats::ToyMCSampler::fPriorNuisance = nullptr

protectedinherited

prior pdf for nuisance parameters

Definition at line 251 of file ToyMCSampler.h.

◆ fProtoData

const RooAbsData* RooStats::ToyMCSampler::fProtoData = nullptr

protectedinherited

in dev

Definition at line 275 of file ToyMCSampler.h.

◆ fReuseNLL

bool RooStats::ToyMCImportanceSampler::fReuseNLL = true

protected

Definition at line 170 of file ToyMCImportanceSampler.h.

◆ fSamplingDistName

std::string RooStats::ToyMCSampler::fSamplingDistName

protectedinherited

name of the model

Definition at line 250 of file ToyMCSampler.h.

◆ fSize

double RooStats::ToyMCSampler::fSize = 0.05

protectedinherited

Definition at line 257 of file ToyMCSampler.h.

◆ fTestStatistics

std::vector<TestStatistic*> RooStats::ToyMCSampler::fTestStatistics

protectedinherited

Definition at line 248 of file ToyMCSampler.h.

◆ fToysInTails

double RooStats::ToyMCSampler::fToysInTails = 0.0

protectedinherited

minimum no of toys in tails for adaptive sampling (taking weights into account, therefore double) Default: 0.0 which means no adaptive sampling

Definition at line 267 of file ToyMCSampler.h.

◆ fToysStrategy

toysStrategies RooStats::ToyMCImportanceSampler::fToysStrategy = EQUALTOYSPERDENSITY

protected

Definition at line 172 of file ToyMCImportanceSampler.h.

◆ fUseMultiGen

bool RooStats::ToyMCSampler::fUseMultiGen = false

protectedinherited

Use PrepareMultiGen?

Definition at line 290 of file ToyMCSampler.h.

The documentation for this class was generated from the following files:

roofit/roostats/inc/RooStats/ToyMCImportanceSampler.h
roofit/roostats/src/ToyMCImportanceSampler.cxx

Public Member Functions

Static Public Member Functions

Protected Member Functions

Protected Attributes

Static Protected Attributes

Constructor & Destructor Documentation

◆ ToyMCImportanceSampler()

◆ ~ToyMCImportanceSampler()

Member Function Documentation

◆ AddImportanceDensity()

◆ AddNullDensity()

◆ AddTestStatistic()

◆ AppendSamplingDistribution()

◆ CheckConfig()

◆ Class()

◆ Class_Name()

◆ Class_Version()

◆ ClearCache()

◆ ConfidenceLevel()

◆ CreateImpDensitiesForOnePOIAdaptively()

◆ CreateNImpDensitiesForOnePOI()

◆ DeclFileName()

◆ EvaluateAllTestStatistics() [1/2]

◆ EvaluateAllTestStatistics() [2/2]

◆ EvaluateTestStatistic() [1/2]

◆ EvaluateTestStatistic() [2/2]

◆ Generate()

◆ GenerateGlobalObservables()

◆ GenerateToyData() [1/7]

◆ GenerateToyData() [2/7]

◆ GenerateToyData() [3/7]

◆ GenerateToyData() [4/7]

◆ GenerateToyData() [5/7]

◆ GenerateToyData() [6/7]

◆ GenerateToyData() [7/7]

◆ GetNToys()

◆ GetSamplingDistName()

◆ GetSamplingDistribution()

◆ GetSamplingDistributions()

◆ GetSamplingDistributionsSingleWorker()

◆ GetTestStatistic() [1/2]

◆ GetTestStatistic() [2/2]

◆ Initialize()

◆ IsA()

◆ nEventsPerToy()

◆ SetAlwaysUseMultiGen()

◆ SetApplyVeto()

◆ SetAsimovNuisancePar()

◆ SetConditionalObservables()

◆ SetConfidenceLevel()

◆ SetDensityToGenerateFromByIndex()

◆ SetEqualNumToysPerDensity()

◆ SetExpectedNuisancePar()

◆ SetExpIncreasingNumToysPerDensity()

◆ SetGenerateAutoBinned()

◆ SetGenerateBinned()

◆ SetGenerateBinnedTag()

◆ SetGlobalObservables()

◆ SetMaxToys()

◆ SetNEventsPerToy()

◆ SetNToys()

◆ SetNuisanceParameters()

◆ SetObservables()

◆ SetParametersForTestStat()

◆ SetPdf()

◆ SetPriorNuisance()

◆ SetProtoData()

◆ SetReuseNLL()

◆ SetSamplingDistName()

◆ SetTestSize()

◆ SetTestStatistic() [1/2]

◆ SetTestStatistic() [2/2]

◆ SetToysBothTails()

◆ SetToysLeftTail()

◆ SetToysRightTail()

◆ SetUseMultiGen()

◆ Streamer()

◆ StreamerNVirtual()

Member Data Documentation

◆ _allVars