A class for handling numerical operations for models. More...

#include <BCIntegrate.h>

Inheritance diagram for BCIntegrate:

Collaboration diagram for BCIntegrate:

Public Types
Enumerators
enum	BCOptimizationMethod { kOptEmpty, kOptSimAnn, kOptMetropolis, kOptMinuit, kOptDefault, NOptMethods }
	An enumerator for the mode finding algorithm. More...

enum	BCIntegrationMethod { kIntEmpty, kIntMonteCarlo, kIntCuba, kIntGrid, kIntLaplace, kIntDefault, NIntMethods }
	An enumerator for integration algorithms. More...

enum	BCMarginalizationMethod { kMargEmpty, kMargMetropolis, kMargMonteCarlo, kMargGrid, kMargDefault, NMargMethods }
	An enumerator for marginalization algorithms. More...

enum	BCSASchedule { kSACauchy, kSABoltzmann, kSACustom, NSAMethods }
	An enumerator for the Simulated Annealing schedule. More...

enum	BCCubaMethod { kCubaVegas, kCubaSuave, kCubaDivonne, kCubaCuhre, kCubaDefault, NCubaMethods }
	An enumerator for Cuba integration methods. More...

Function pointer types
typedef void(BCIntegrate::*	tRandomizer) (std::vector< double > &) const
	A pointer for a function that chooses a next random point.

typedef double(BCIntegrate::*	tEvaluator) (std::vector< double > &, const std::vector< double > &, bool &)
	A pointer for a function that evaluates at a point.

typedef void(*	tIntegralUpdater) (const std::vector< double > &, const int &, double &, double &)
	A pointer for a function that updates the integral and absolute precision.

Public Types inherited from BCEngineMCMC
enum	Precision { kLow, kQuick, kMedium, kHigh, kVeryHigh }
	An enumerator for the status of a test. More...

enum	Phase { kPreRun = -1, kUnsetPhase = 0, kMainRun = +1 }
	An enumerator for the phase of the Markov chain. More...

enum	InitialPositionScheme { kInitCenter = 0, kInitRandomUniform = 1, kInitUserDefined = 2, kInitRandomPrior = 3 }
	An enumerator for markov-chain position initialization. More...

Public Member Functions
Constructors and destructors
	BCIntegrate (const std::string &name="model")
	Default constructor.

	BCIntegrate (const std::string &filename, const std::string &name, bool loadObservables=true)
	Read in MCMC constructor. More...

	BCIntegrate (const BCIntegrate &other)
	Copy constructor.

BCIntegrate &	operator= (const BCIntegrate &)
	Copy-assignment operator.

virtual	~BCIntegrate ()
	Destructor.

Member functions (get)
double	GetIntegral () const

BCIntegrate::BCOptimizationMethod	GetOptimizationMethod () const

BCIntegrate::BCIntegrationMethod	GetIntegrationMethod () const

BCIntegrate::BCMarginalizationMethod	GetMarginalizationMethod () const

BCIntegrate::BCSASchedule	GetSASchedule () const

void	GetRandomVectorInParameterSpace (std::vector< double > &x) const
	Fills a vector of random numbers x[i] between fMin[i] and fMax[i] into a vector. More...

double	GetRandomPoint (std::vector< double > &x)
	Fills a vector of (flat) random numbers in the limits of the parameters and returns the probability at that point. More...

int	GetNIterationsMin () const

int	GetNIterationsMax () const

int	GetNIterationsPrecisionCheck () const

int	GetNIterations () const

double	GetRelativePrecision () const

double	GetAbsolutePrecision () const

BCCubaMethod	GetCubaIntegrationMethod () const

const BCCubaOptions::Vegas &	GetCubaVegasOptions () const

const BCCubaOptions::Suave &	GetCubaSuaveOptions () const

const BCCubaOptions::Divonne &	GetCubaDivonneOptions () const

const BCCubaOptions::Cuhre &	GetCubaCuhreOptions () const

TH1 *	GetSlice (std::vector< unsigned > indices, unsigned &nIterations, double &log_max_val, const std::vector< double > parameters=std::vector< double >(0), int nbins=0, bool normalize=true)
	Returns a one-dimensional slice of the pdf at the point and along a specific direction. More...

TH1 *	GetSlice (const std::string &name, unsigned &nIterations, double &log_max_val, const std::vector< double > parameters=std::vector< double >(0), int nbins=0, bool normalize=true)
	Returns a one-dimensional slice of the pdf at the point and along a specific direction. More...

TH1 *	GetSlice (unsigned index, unsigned &nIterations, double &log_max_val, const std::vector< double > parameters=std::vector< double >(0), int nbins=0, bool normalize=true)
	Returns a one-dimensional slice of the pdf at the point and along a specific direction. More...

TH2 *	GetSlice (const std::string &name1, const std::string &name2, unsigned &nIterations, double &log_max_val, const std::vector< double > parameters=std::vector< double >(0), int nbins=0, bool normalize=true)
	Returns a two-dimensional slice of the pdf at the point and along two specified directions. More...

TH2 *	GetSlice (unsigned index1, unsigned index2, unsigned &nIterations, double &log_max_val, const std::vector< double > parameters=std::vector< double >(0), int nbins=0, bool normalize=true)
	Returns a two-dimensional slice of the pdf at the point and along two specified directions. More...

double	GetError () const

TMinuitMinimizer &	GetMinuit ()

double	GetSAT0 () const
	Returns the Simulated Annealing starting temperature. More...

double	GetSATmin () const
	Returns the Simulated Annealing threshhold temperature. More...

virtual const std::vector< double > &	GetBestFitParameters () const

const std::vector< double > &	GetBestFitParameterErrors () const
	Returns the set of errors on the values of the parameters at the mode.

double	GetLogMaximum () const
	Returns the posterior at the mode. More...

Member functions (set)
void	SetFlagIgnorePrevOptimization (bool flag)

void	SetOptimizationMethod (BCIntegrate::BCOptimizationMethod method)

void	SetIntegrationMethod (BCIntegrate::BCIntegrationMethod method)

void	SetMarginalizationMethod (BCIntegrate::BCMarginalizationMethod method)

void	SetSASchedule (BCIntegrate::BCSASchedule schedule)

void	SetNIterationsMin (int niterations)

void	SetNIterationsMax (int niterations)

void	SetNIterationsPrecisionCheck (int niterations)

void	SetRelativePrecision (double relprecision)

void	SetAbsolutePrecision (double absprecision)
	Set absolute precision of the numerical integation.

void	SetCubaIntegrationMethod (BCCubaMethod type)
	Set Cuba integration method.

void	SetCubaOptions (const BCCubaOptions::Vegas &options)
	Set options for CUBA's Vegas. More...

void	SetCubaOptions (const BCCubaOptions::Suave &options)
	Set options for CUBA's Suave. More...

void	SetCubaOptions (const BCCubaOptions::Divonne &options)
	Set options for CUBA's Divonne. More...

void	SetCubaOptions (const BCCubaOptions::Cuhre &options)
	Set options for CUBA's Cuhre. More...

void	SetSAT0 (double T0)
	Set starting temperature for Simulated Annealing. More...

void	SetSATmin (double Tmin)
	Set threshold temperature for Simulated Annealing. More...

Public Member Functions inherited from BCEngineMCMC
	BCEngineMCMC (const std::string &name="model")
	Default constructor. More...

	BCEngineMCMC (const BCEngineMCMC &enginemcmc)
	Copy constructor. More...

	BCEngineMCMC (const std::string &filename, const std::string &name, bool loadObservables=true)
	Read in MCMC constructor. More...

BCEngineMCMC &	operator= (const BCEngineMCMC &)
	Copy-assignment operator.

virtual	~BCEngineMCMC ()
	Destructor. More...

const std::string &	GetName () const

const std::string &	GetSafeName () const

unsigned	GetNChains () const

unsigned	GetNLag () const

int	GetCurrentIteration () const

unsigned	GetCurrentChain () const

int	GetNIterationsConvergenceGlobal () const

unsigned	GetNIterationsPreRun () const

unsigned	GetNIterationsPreRunMin () const

unsigned	GetNIterationsPreRunMax () const

unsigned	GetNIterationsRun () const

unsigned	GetNIterationsPreRunCheck () const

unsigned	GetPreRunCheckClear ()

double	GetMinimumEfficiency () const

double	GetMaximumEfficiency () const

double	GetScaleFactorLowerLimit () const

double	GetScaleFactorUpperLimit () const

const std::vector< std::vector< double > > &	GetScaleFactors () const

const ChainState &	GetChainState (unsigned c) const

const std::vector< double > &	Getx (unsigned c) const

double	Getx (unsigned c, unsigned p) const

double	GetLogProbx (unsigned c) const

BCEngineMCMC::Phase	GetPhase () const

BCEngineMCMC::InitialPositionScheme	GetInitialPositionScheme () const

unsigned	GetInitialPositionAttemptLimit () const

bool	GetProposeMultivariate () const

double	GetProposalFunctionDof () const

unsigned	GetMultivariateCovarianceUpdates () const

double	GetMultivariateCovarianceUpdateLambda () const

double	GetMultivariateEpsilon () const

double	GetMultivariateScaleMultiplier () const

double	GetRValueParametersCriterion () const

const std::vector< double > &	GetRValueParameters () const

double	GetRValueParameters (unsigned index) const

bool	GetCorrectRValueForSamplingVariability () const
	Flag for correcting convergence checking for initial sampling variability. More...

bool	GetFlagRun () const

TTree *	GetMarkovChainTree () const
	Retrieve the tree containing the Markov chain. More...

TTree *	GetParameterTree () const
	Retrieve the tree containing the parameter information. More...

TFile *	GetOutputFile () const
	Retrieve output file for MCMC. More...

const BCEngineMCMC::Statistics &	GetStatistics () const
	Get combined statistics for all chains. More...

const BCEngineMCMC::Statistics &	GetStatistics (unsigned c) const
	Get MCMC statistics for one chain. More...

const std::vector< BCEngineMCMC::Statistics > &	GetStatisticsVector () const
	Get vector of MCMC statistics for each chain separately. More...

bool	GetRescaleHistogramRangesAfterPreRun () const

double	GetHistogramRescalePadding () const

virtual std::vector< unsigned >	GetH1DPrintOrder () const

virtual std::vector< std::pair< unsigned, unsigned > >	GetH2DPrintOrder () const

bool	MarginalizedHistogramExists (unsigned index) const

bool	MarginalizedHistogramExists (unsigned index1, unsigned index2) const

TH1 *	GetMarginalizedHistogram (const std::string &name) const
	Obtain the individual marginalized distributions with respect to one parameter as a ROOT TH1. More...

TH1 *	GetMarginalizedHistogram (unsigned index) const
	Obtain the individual marginalized distributions with respect to one parameter as a ROOT TH1. More...

TH2 *	GetMarginalizedHistogram (const std::string &name1, const std::string &name2) const
	Obtain the individual marginalized distributions with respect to two parameters as a ROOT TH2. More...

TH2 *	GetMarginalizedHistogram (unsigned index1, unsigned index2) const
	Obtain the individual marginalized distributions with respect to two parameters as a ROOT TH2. More...

BCH1D	GetMarginalized (const std::string &name) const
	Obtain the individual marginalized distributions with respect to one parameter. More...

BCH1D	GetMarginalized (unsigned index) const
	Obtain the individual marginalized distributions with respect to one parameter. More...

BCH2D	GetMarginalized (const std::string &name1, const std::string &name2) const
	Obtain the individual marginalized distributions with respect to two parameters. More...

BCH2D	GetMarginalized (unsigned index1, unsigned index2) const
	Obtain the individual marginalized distributions with respect to two parameters. More...

unsigned	GetMaximumParameterNameLength (bool observables=true) const

BCVariable &	GetVariable (unsigned index)

const BCVariable &	GetVariable (unsigned index) const

unsigned	GetNVariables () const

BCParameterSet &	GetParameters ()

const BCParameterSet &	GetParameters () const

BCParameter &	GetParameter (unsigned index)

const BCParameter &	GetParameter (unsigned index) const

BCParameter &	GetParameter (const std::string &name)

const BCParameter &	GetParameter (const std::string &name) const

unsigned	GetNParameters () const

unsigned	GetNFixedParameters () const

unsigned	GetNFreeParameters () const

BCObservableSet &	GetObservables ()

const BCObservableSet &	GetObservables () const

BCObservable &	GetObservable (unsigned index)

const BCObservable &	GetObservable (unsigned index) const

BCObservable &	GetObservable (const std::string &name)

const BCObservable &	GetObservable (const std::string &name) const

unsigned	GetNObservables () const

const std::vector< double > &	GetLocalModes (bool force_recalculation=false)

bool	GetReuseObservables () const

BCH1D &	GetBCH1DdrawingOptions ()

BCH2D &	GetBCH2DdrawingOptions ()

void	SetName (const std::string &name)
	Sets the name of the engine. More...

void	SetScaleFactorLowerLimit (double l)
	Set scale factor lower limit.

void	SetScaleFactorUpperLimit (double l)
	Set scale factor upper limit.

void	SetInitialScaleFactors (const std::vector< double > &scale)
	Set the initial scale factors for the factorized proposal function. More...

void	SetNChains (unsigned n)
	Sets the number of Markov chains which are run in parallel. More...

void	SetNLag (unsigned n)
	Sets the lag of the Markov chains.

void	SetNIterationsPreRunMax (unsigned n)
	Sets the maximum number of iterations in the pre-run. More...

void	SetNIterationsRun (unsigned n)
	Sets the number of iterations. More...

void	SetNIterationsPreRunMin (unsigned n)
	Sets the minimum number of iterations in the pre-run.

void	SetNIterationsPreRunCheck (unsigned n)
	Sets the number of iterations between scale adjustments and convergence checks in the pre-run. More...

void	SetPreRunCheckClear (unsigned n)
	Sets the number of prerun checks to make inbetween statistics clearing. More...

void	SetMinimumEfficiency (double efficiency)
	Sets the minimum efficiency required for a chain. More...

void	SetMaximumEfficiency (double efficiency)
	Sets the maximum efficiency required for a chain. More...

void	SetRandomSeed (unsigned seed)
	Set the random number seed.

void	SetInitialPositions (const std::vector< double > &x0s)
	Sets the initial positions for all chains. More...

void	SetInitialPositions (const std::vector< std::vector< double > > &x0s)
	Sets the initial positions for all chains. More...

void	SetInitialPositionScheme (BCEngineMCMC::InitialPositionScheme scheme)
	Sets flag which defines initial position. More...

void	SetInitialPositionAttemptLimit (unsigned n)
	Sets maximum number of attempts to find a valid initial position. More...

void	SetProposeMultivariate (bool flag)
	Set `flag` to `true` to turn on the multivariate proposal for MCMC based on (Haario et al., 2001) where the covariance is learned from the prerun. More...

void	SetProposalFunctionDof (double dof=1)
	Set the degree of freedom of the proposal function for MCMC. More...

void	SetMultivariateCovarianceUpdateLambda (double l)
	Set weighting for multivariate proposal function covariance update. More...

void	SetMultivariateEpsilon (double epsilon)
	Sets multivariate-proposal-function cholesky-decomposition nudge. More...

void	SetMultivariateScaleMultiplier (double s)
	Sets multivariate-proposal-function scale multiplier. More...

void	SetFlagFillHistograms (bool flag)
	Sets whether to fill histograms. More...

void	SetFlagFillHistograms (bool flag_1d, bool flag_2d)
	Sets the whether to fill histograms. More...

void	SetFillHistogramParPar (unsigned x, unsigned y, bool flag=true)
	Sets whether to fill particular H2 histogram: par(y) vs. More...

void	SetFillHistogramParPar (const std::string &x, const std::string &y, bool flag=true)
	Sets whether to fill particular H2 histogram: par(y) vs. More...

void	SetFillHistogramParObs (unsigned x, unsigned y, bool flag=true)
	Sets whether to fill particular H2 histogram: obs(y) vs. More...

void	SetFillHistogramParObs (const std::string &x, const std::string &y, bool flag=true)
	Sets whether to fill particular H2 histogram: obs(y) vs. More...

void	SetFillHistogramObsObs (unsigned x, unsigned y, bool flag=true)
	Sets whether to fill particular H2 histogram: obs(y) vs. More...

void	SetFillHistogramObsObs (const std::string &x, const std::string &y, bool flag=true)
	Sets whether to fill particular H2 histogram: obs(y) vs. More...

void	SetFillHistogramObsPar (unsigned x, unsigned y, bool flag=true)
	Sets whether to fill particular H2 histogram: par(y) vs. More...

void	SetFillHistogramObsPar (const std::string &x, const std::string &y, bool flag=true)
	Sets whether to fill particular H2 histogram: par(y) vs. More...

void	SetFlagPreRun (bool flag)
	Set if a (new) prerun should be performed. More...

void	SetRValueParametersCriterion (double r)
	Sets the parameter R-value criterion for convergence of all chains.

void	SetCorrectRValueForSamplingVariability (bool flag=true)
	Set flag to correct convergence checking for initial sampling variability. More...

void	SetPrecision (BCEngineMCMC::Precision precision)
	Set the precision for the MCMC run. More...

void	SetPrecision (const BCEngineMCMC *other)
	Copy precision for the MCMC run from other model. More...

void	SetPrecision (const BCEngineMCMC &other)
	Copy precision for the MCMC run from other model. More...

void	SetNbins (unsigned int nbins)
	Set the number of bins for the marginalized distribution of all parameters. More...

void	SetReuseObservables (bool flag)

void	SetRescaleHistogramRangesAfterPreRun (bool flag=true)
	Set flag for rescaling histogram ranges after pre-run. More...

void	SetHistogramRescalingPadding (double factor)
	Set enlargement factor of range for when rescaling. More...

void	WriteMarkovChain (bool flag)
	Turn on/off writing of Markov chain to root file. More...

void	WriteMarkovChainRun (bool flag)
	Turn on/off writing of Markov chain to root file during run. More...

void	WriteMarkovChainPreRun (bool flag)
	Turn on/off writing of Markov chain to root file during prerun. More...

void	WriteMarkovChain (const std::string &filename, const std::string &option, bool flag_run=true, bool flag_prerun=true)
	Turn on writing of Markov chain to root file. More...

void	SetPriorConstant (unsigned index)

void	SetPriorConstant (const std::string &name)

void	SetPrior (unsigned index, TF1 &f, bool logL=true)

void	SetPrior (const std::string &name, TF1 &f, bool logL=true)

void	SetPriorDelta (unsigned index, double value)

void	SetPriorDelta (const std::string &name, double value)

void	SetPriorGauss (unsigned index, double mean, double sigma)

void	SetPriorGauss (const std::string &name, double mean, double sigma)

void	SetPriorGauss (unsigned index, double mode, double sigma_below, double sigma_above)

void	SetPriorGauss (const std::string &name, double mode, double sigma_below, double sigma_above)

void	SetPrior (unsigned index, TH1 &h, bool interpolate=false)

void	SetPrior (const std::string &name, TH1 &h, bool interpolate=false)

void	SetPriorConstantAll ()

void	WriteMarginalizedDistributions (const std::string &filename, const std::string &option, bool closeExistingFile=false)
	Write marginalization histograms to file. More...

virtual void	PrintSummary () const
	Prints a summary to the logs. More...

void	PrintParameters (const std::vector< double > &P, void(*output)(const std::string &)=BCLog::OutSummary) const
	Print parameters. More...

unsigned	PrintAllMarginalized (const std::string &filename, unsigned hdiv=1, unsigned vdiv=1) const
	Print all marginalizations. More...

unsigned	PrintParameterPlot (const std::string &filename, int npar=10, double interval_content=68e-2, std::vector< double > quantile_values=std::vector< double >(0), bool rescale_ranges=true) const
	Print a summary plot for the parameters and user-defined observables. More...

bool	DrawParameterPlot (unsigned i0, unsigned npar=0, double interval_content=68e-2, std::vector< double > quantile_values=std::vector< double >(0), bool rescale_ranges=true) const
	Draw a summary plot for the parameters in the range provided to current pad. More...

bool	PrintCorrelationMatrix (const std::string &filename="matrix.pdf") const
	Print a correlation matrix for the parameters. More...

bool	PrintCorrelationPlot (const std::string &filename="correlation.pdf", bool include_observables=true) const
	Print a correlation plot for the parameters. More...

bool	PrintParameterLatex (const std::string &filename) const
	Print a LaTeX table of the parameters. More...

virtual void	CreateHistograms (bool rescale_ranges=false)
	Create histograms from parameter and observable sets. More...

virtual void	InitializeMarkovChainTree (bool replacetree=false, bool replacefile=false)
	Initialize the trees containing the Markov chains and parameter info. More...

virtual bool	AddParameter (const std::string &name, double min, double max, const std::string &latexname="", const std::string &unitstring="")

virtual bool	AddParameter (BCParameter &parameter)

virtual bool	AddObservable (const std::string &name, double min, double max, const std::string &latexname="", const std::string &unitstring="")

virtual bool	AddObservable (BCObservable &obs)

virtual void	EvaluateObservables ()
	Evaluates user-defined observables at current state of all chains and stores results in fMCMCState.

virtual void	EvaluateObservables (unsigned chain)
	Evaluates user-defined observables at current state of chain and stores results in fMCMCState. More...

virtual void	CalculateObservables (const std::vector< double > &pars)
	Evaluates user-defined observables. More...

virtual double	ProposalFunction (unsigned ichain, unsigned ipar)
	The default proposal function is a Breit-Wigner random walk. More...

bool	GetProposalPointMetropolis (unsigned chain, std::vector< double > &x)
	Return a proposal point for the Metropolis algorithm. More...

bool	GetProposalPointMetropolis (unsigned chain, unsigned parameter, std::vector< double > &x)
	Return a proposal point for the Metropolis algorithm. More...

bool	GetNewPointMetropolis ()
	Generate a new point using the Metropolis algorithm for all chains. More...

bool	GetNewPointMetropolis (unsigned chain)
	Generate a new point using the Metropolis algorithm for one chain. More...

bool	GetNewPointMetropolis (unsigned chain, unsigned parameter)
	Generate a new point using the Metropolis algorithm for one chain, varying only one parameter's value. More...

bool	AcceptOrRejectPoint (unsigned chain, unsigned parameter)
	Accept or rejects a point for a chain and updates efficiency. More...

void	InChainFillHistograms (const ChainState &cs)
	Fill marginalized distributions from a chain state.

void	InChainFillHistograms ()
	Fill marginalized distributions from all chain states.

void	InChainFillTree (const ChainState &cs, unsigned chain_number)
	Write a chain state to the tree.

void	InChainFillTree ()
	Write all chain states to the tree.

bool	Metropolis ()
	Runs Metropolis algorithm. More...

bool	MetropolisPreRun ()
	Runs a pre run for the Metropolis algorithm. More...

void	MCMCInitialize ()
	Resets all containers used in MCMC and initializes starting points. More...

virtual void	MCMCUserInitialize ()
	User hook called from MCMCInitialize(). More...

virtual void	MCMCUserIterationInterface ()
	Interface allowing to execute arbitrary code for each iteration of the MCMC while running the chains after applying the lag but before filling histograms or the output tree. More...

virtual void	MCMCCurrentPointInterface (const std::vector< double > &point, int ichain, bool accepted)
	Interface allowing to execute arbitrary code for each new point of the MCMC whether it is accepted or not. More...

void	LoadParametersFromTree (TTree *partree, bool loadObservables=true)
	Load parameters and observables from tree. More...

void	LoadMCMCParameters (TTree &partree)
	Load MCMC parameters from parameter tree: nchains, proposal function type, scales. More...

virtual bool	ParameterTreeMatchesModel (TTree *partree, bool checkObservables=true)
	Check parameter tree against model. More...

void	LoadMCMC (const std::string &filename, std::string mcmcTreeName="", std::string parameterTreeName="", bool loadObservables=true)
	Load previous MCMC run. More...

void	LoadMCMC (TTree mcmcTree, TTree parTree, bool loadObservables=true)
	Load previous MCMC run. More...

bool	ValidMCMCTree (TTree *tree, bool checkObservables=true) const
	Check tree structure for MCMC tree. More...

bool	ValidParameterTree (TTree *tree) const
	Check tree structure for parameter tree. More...

void	CloseOutputFile ()
	Close the root output file. More...

virtual void	Remarginalize (bool autorange=true)
	Marginalize from TTree. More...

void	PrepareToContinueMarginalization (const std::string &filename, const std::string &mcmcTreeName="", const std::string &parameterTreeName="", bool loadObservables=true, bool autorange=true)
	Continue the marginalization already stored in another file. More...

virtual bool	UpdateMultivariateProposalFunctionCovariances (double a)
	Update multivariate proposal function covariances. More...

virtual bool	UpdateMultivariateProposalFunctionCovariances ()
	Update multivariate proposal function covariances. More...

void	CalculateCholeskyDecompositions ()
	Calculate Cholesky decompositions needed for multivariate proposal function. More...

void	UpdateChainIndex (int chain)
	Keep track of which chain is currently computed (within a thread). More...

Protected Member Functions
virtual std::string	GetBestFitSummary (unsigned i) const
	Get string summarizing best fit for single variable. More...

unsigned	IntegrationOutputFrequency () const
	Determine frequency of output during integration.

void	LogOutputAtEndOfIntegration (double integral, double absprecision, double relprecision, int nIterations)
	Helper method to output at end of integration. More...

void	LogOutputAtIntegrationStatusUpdate (BCIntegrationMethod type, double integral, double absprecision, int nIterations)
	Helper method to output integration status. More...

void	LogOutputAtStartOfIntegration (BCIntegrationMethod type, BCCubaMethod cubatype)
	Helper method to output at beginning of integration. More...

virtual void	PrintBestFitSummary () const
	Print best fit to log.

virtual void	PrintMarginalizationSummary () const
	Print marginalization to log. More...

Protected Member Functions inherited from BCEngineMCMC
virtual void	PrintModelSummary () const
	Print model summary to log. More...

void	SetFillHistogram (int x, int y, bool flag)
	Set whether to fill 2D histogram y vs x: positive indices for parameters; negative for observables, starting at -1 and going more negative—observable index = -(index+1). More...

unsigned	UpdateFrequency (unsigned N) const
	return appropriate update interval More...

void	UpdateParameterTree ()
	Update Paramater TTree with scales and efficiencies. More...

Protected Attributes
bool	fFlagIgnorePrevOptimization
	Flag for ignoring older results of optimization.

bool	fFlagMarginalized
	flag indicating if the model was marginalized

double	fSALogProb
	Log probability of current simulated annealing iteration. More...

int	fSANIterations
	Number of iterations for simualted annealing. More...

double	fSAT0
	Starting temperature for Simulated Annealing.

double	fSATemperature
	Current temperature of simulated annealing algorithm. More...

double	fSATmin
	Minimal/Threshold temperature for Simulated Annealing.

std::vector< double >	fSAx
	Current simulated annealing parameter point. More...

Protected Attributes inherited from BCEngineMCMC
BCH1D	fBCH1DdrawingOptions
	A BCH1D (with no histogram) for storing BCH1D drawing options. More...

BCH2D	fBCH2DdrawingOptions
	A BCH2D (with no histogram) for storing BCH2D drawing options. More...

bool	fCorrectRValueForSamplingVariability
	flag for correcting R value for initial sampling variability. More...

std::vector< TH1 * >	fH1Marginalized
	Vector of 1D marginalized distributions.

std::vector< std::vector< TH2 * > >	fH2Marginalized
	Vector of 2D marginalized distributions. More...

double	fHistogramRescalePadding
	factor for enlarging range of histograms when rescaling. More...

unsigned	fInitialPositionAttemptLimit
	Maximum number of attempts to make to set the initial position. More...

BCEngineMCMC::InitialPositionScheme	fInitialPositionScheme
	Variable which defines the initial position. More...

std::vector< double >	fLocalModes
	Vector of local modes. More...

int	fMCMCCurrentIteration
	The current iteration number. More...

double	fMCMCEfficiencyMax
	The maximum allowed efficiency for MCMC.

double	fMCMCEfficiencyMin
	The minimum required efficiency for MCMC.

bool	fMCMCFlagWriteChainToFile
	Flag to write Markov chains to file.

bool	fMCMCFlagWritePreRunToFile
	Flag to write pre run to file.

std::vector< std::vector< double > >	fMCMCInitialPosition
	The intial position of each Markov chain. More...

std::vector< double >	fMCMCInitialScaleFactors
	User-provided initial values of the scale factors of the factorized proposal function. More...

unsigned	fMCMCNChains
	Number of Markov chains ran in parallel.

int	fMCMCNIterationsConvergenceGlobal
	Number of iterations needed for all chains to convergence simultaneously.

unsigned	fMCMCNIterationsPreRunCheck
	Number of iterations between scale adjustments and convergence checks in pre-run. More...

unsigned	fMCMCNIterationsPreRunMax
	Maximum number of iterations for a Markov chain prerun.

unsigned	fMCMCNIterationsPreRunMin
	Minimum number of iterations for the pre-run.

unsigned	fMCMCNIterationsRun
	Number of iterations for a Markov chain run.

unsigned	fMCMCNLag
	The lag for the Markov Chain.

TFile *	fMCMCOutputFile
	Output file for for writing MCMC Tree. More...

std::string	fMCMCOutputFilename
	Output filename for for writing MCMC Tree. More...

std::string	fMCMCOutputFileOption
	Output file open option for for writing MCMC Tree. More...

BCEngineMCMC::Phase	fMCMCPhase
	The phase of the run. More...

unsigned	fMCMCPreRunCheckClear
	Number of iterations between clearing of convergence stats in pre-run. More...

double	fMCMCProposalFunctionDof
	Degree of freedom of Student's t proposal. More...

std::vector< std::vector< double > >	fMCMCProposalFunctionScaleFactor
	Scale factors for proposal functions. More...

bool	fMCMCProposeMultivariate
	Flag for using multivariate proposal function. More...

std::vector< double >	fMCMCRValueParameters
	The R-values for each parameter.

double	fMCMCRValueParametersCriterion
	The R-value criterion for convergence of parameters.

double	fMCMCScaleFactorLowerLimit
	Lower limit for scale factors.

double	fMCMCScaleFactorUpperLimit
	Upper limit for scale factors.

std::vector< ChainState >	fMCMCStates
	The current states of each Markov chain. More...

std::vector< BCEngineMCMC::Statistics >	fMCMCStatistics
	Statistics for each Markov chain. More...

BCEngineMCMC::Statistics	fMCMCStatistics_AllChains
	Statistics across all Markov chains. More...

TTree *	fMCMCTree
	The tree containing the Markov chains. More...

unsigned int	fMCMCTree_Chain
	Chain number for storing into tree.

ChainState	fMCMCTree_State
	MC state object for storing into tree.

bool	fMCMCTreeLoaded
	flag for whether MCMC Tree successfully loaded. More...

bool	fMCMCTreeReuseObservables
	flag for whether to reuse MCMC Tree's observables. More...

double	fMultivariateCovarianceUpdateLambda
	weighting parameter for multivariate-proposal-function covariance update. More...

unsigned	fMultivariateCovarianceUpdates
	Number of multivariate-proposal-function covariance updates performed. More...

double	fMultivariateEpsilon
	multivariate-proposal-function cholesky-decomposition nudge. More...

std::vector< TMatrixD >	fMultivariateProposalFunctionCholeskyDecomposition
	Cholesky decompositions for multivariate proposal function. More...

std::vector< TMatrixDSym >	fMultivariateProposalFunctionCovariance
	Covariance matrices used in multivariate proposal functions. More...

double	fMultivariateScaleMultiplier
	factor to multiply or divide scale factors by in adjusting multivariate-proposal-function scales. More...

std::string	fName
	Name of the engine. More...

BCAux::BCTrash< TObject >	fObjectTrash
	Storage for plot objects with proper clean-up.

BCObservableSet	fObservables
	User-calculated Observables Set.

BCParameterSet	fParameters
	Parameter settings.

TTree *	fParameterTree
	The tree containing the parameter information. More...

TRandom3	fRandom
	Random number generator.

std::vector< std::pair< int, int > >	fRequestedH2
	Vector of pairs of indices for which 2D histograms should be stored. More...

bool	fRescaleHistogramRangesAfterPreRun
	flag for rescaling of histograms after pre-run. More...

std::string	fSafeName
	Safe name of the engine for use in naming ROOT objects. More...

Member functions (miscellaneous methods)
virtual double	Eval (const std::vector< double > &x)=0
	Evaluate the unnormalized probability at a point in parameter space. More...

virtual double	LogEval (const std::vector< double > &x)
	Evaluate the natural logarithm of the Eval function. More...

double	Normalize ()
	Performs integration. More...

double	Integrate (BCIntegrationMethod intmethod)
	Does the integration over the un-normalized probability. More...

double	Integrate ()
	Perform the integration. More...

double	Integrate (BCIntegrationMethod type, tRandomizer randomizer, tEvaluator evaluator, tIntegralUpdater updater, std::vector< double > &sums)
	Does the integration over the un-normalized probability. More...

double	EvaluatorMC (std::vector< double > &sums, const std::vector< double > &point, bool &accepted)
	Evaluates integrator.

int	MarginalizeAll ()
	Marginalize all probabilities wrt. More...

int	MarginalizeAll (BCMarginalizationMethod margmethod)
	Marginalize all probabilities wrt. More...

virtual void	MarginalizePreprocess ()
	Method executed for before marginalization. More...

virtual void	MarginalizePostprocess ()
	Method executed after marginalization. More...

std::vector< double >	FindMode (std::vector< double > start=std::vector< double >())
	Do the mode finding using a method set via SetOptimizationMethod. More...

std::vector< double >	FindMode (BCIntegrate::BCOptimizationMethod optmethod, std::vector< double > start=std::vector< double >())
	Find mode using a specific method. More...

double	SATemperature (double t) const
	Temperature annealing schedule for use with Simulated Annealing. More...

double	SATemperatureBoltzmann (double t) const
	Temperature annealing schedule for use with Simulated Annealing. More...

double	SATemperatureCauchy (double t) const
	Temperature annealing schedule for use with Simulated Annealing. More...

virtual double	SATemperatureCustom (double t) const
	Temperature annealing schedule for use with Simulated Annealing. More...

std::vector< double >	GetProposalPointSA (const std::vector< double > &x, int t) const
	Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm. More...

std::vector< double >	GetProposalPointSABoltzmann (const std::vector< double > &x, int t) const
	Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm. More...

std::vector< double >	GetProposalPointSACauchy (const std::vector< double > &x, int t) const
	Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm. More...

virtual std::vector< double >	GetProposalPointSACustom (const std::vector< double > &x, int t) const
	Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm. More...

std::vector< double >	SAHelperGetRandomPointOnHypersphere () const
	Generates a uniform distributed random point on the surface of a fNvar-dimensional Hypersphere. More...

double	SAHelperGetRadialCauchy () const
	Generates the radial part of a n-dimensional Cauchy distribution. More...

double	SAHelperSinusToNIntegral (int dim, double theta) const
	Returns the Integral of sin^dim from 0 to theta. More...

virtual void	ResetResults ()
	Reset all information on the best-fit parameters. More...

std::string	DumpIntegrationMethod (BCIntegrationMethod type) const
	Return string with the name for a given integration type. More...

std::string	DumpCurrentIntegrationMethod () const
	Return string with the name for the currently set integration type. More...

std::string	DumpUsedIntegrationMethod () const
	Return string with the name for the previously used integration type. More...

std::string	DumpMarginalizationMethod (BCMarginalizationMethod type) const
	Return string with the name for a given marginalization type. More...

std::string	DumpCurrentMarginalizationMethod () const
	Return string with the name for the currently set marginalization type. More...

std::string	DumpUsedMarginalizationMethod () const
	Return string with the name for the marginalization type used. More...

std::string	DumpOptimizationMethod (BCOptimizationMethod type) const
	Return string with the name for a given optimization type. More...

std::string	DumpCurrentOptimizationMethod () const
	Return string with the name for the currently set optimization type. More...

std::string	DumpUsedOptimizationMethod () const
	Return string with the name for the optimization type used to find the current mode. More...

std::string	DumpCubaIntegrationMethod (BCCubaMethod type) const
	Return string with the name for a given Cuba integration type. More...

std::string	DumpCubaIntegrationMethod () const
	Return string with the name for the currently set Cuba integration type. More...

void	SetBestFitParameters (const std::vector< double > &x)
	Set best fit parameters values. More...

void	SetBestFitParameters (const std::vector< double > &x, const double &new_value, double &old_value)

bool	CheckMarginalizationAvailability (BCMarginalizationMethod type)
	Check availability of integration routine for marginalization. More...

bool	CheckMarginalizationIndices (TH1 *hist, const std::vector< unsigned > &index)
	Check that indices of parameters to marginalize w/r/t are correct.

double	IntegrateLaplace ()
	Integrate using the Laplace approximation. More...

static void	IntegralUpdaterMC (const std::vector< double > &sums, const int &nIterations, double &integral, double &absprecision)
	Updates info about integrator.

Additional Inherited Members
Static Public Member Functions inherited from BCEngineMCMC
static double	RValue (const std::vector< double > &means, const std::vector< double > &variances, unsigned n, bool correctForSamplingVariability=true)
	Calculate R value of set of batches of samples—represented by their means and variances, all batches containing the same number of samples—according to Brooks & Gelman, "General Methods for Monitoring Convergence of Iterative Simulations," (1988) More...

Detailed Description

A class for handling numerical operations for models.

Author: Daniel Kollar; Kevin Kröninger

Version: 1.0

Date: 08.2008

This is a base class for a model class. It contains numerical methods to carry out the integration, marginalization, peak finding etc.

Definition at line 143 of file BCIntegrate.h.

Member Enumeration Documentation

enum BCIntegrate::BCCubaMethod

An enumerator for Cuba integration methods.

Enumerator
kCubaVegas	Vegas.
kCubaSuave	Suave.
kCubaDivonne	Divonne.
kCubaCuhre	Cuhre.
kCubaDefault	Default.
NCubaMethods	number of available CUBA methods

Definition at line 196 of file BCIntegrate.h.

enum BCIntegrate::BCIntegrationMethod

An enumerator for integration algorithms.

Enumerator
kIntEmpty	No integration method set.
kIntMonteCarlo	Sample mean method.
kIntCuba	Use CUBA interface.
kIntGrid	Integration by gridding of parameter space.
kIntLaplace	Laplace approximation.
kIntDefault	Default.
NIntMethods	number of available integration methods

Definition at line 164 of file BCIntegrate.h.

enum BCIntegrate::BCMarginalizationMethod

An enumerator for marginalization algorithms.

Enumerator
kMargEmpty	No marginalization method set.
kMargMetropolis	Metropolis Hastings.
kMargMonteCarlo	Sample mean Monte Carlo.
kMargGrid	Marginalization by gridding of parameter space.
kMargDefault	Default.
NMargMethods	number of available marginalization methods

Definition at line 176 of file BCIntegrate.h.

enum BCIntegrate::BCOptimizationMethod

An enumerator for the mode finding algorithm.

Enumerator
kOptEmpty	No optimization method set.
kOptSimAnn	Simulated annealing.
kOptMetropolis	Metropolis Hastings.
kOptMinuit	ROOT's Minuit.
kOptDefault	Default.
NOptMethods	number of available optimization methods

Definition at line 153 of file BCIntegrate.h.

enum BCIntegrate::BCSASchedule

An enumerator for the Simulated Annealing schedule.

Enumerator
kSACauchy	Cauchy scheduler.
kSABoltzmann	Boltzman scheduler.
kSACustom	Custom scheduler.
NSAMethods	number of available schedulers

Definition at line 187 of file BCIntegrate.h.

Constructor & Destructor Documentation

BCIntegrate::BCIntegrate	(	const std::string &	filename,
		const std::string &	name,
		bool	loadObservables = `true`
	)

Read in MCMC constructor.

Parameters

filename	Path of file holding model.
name	Name of model (file should contain TTree's [name]_mcmc and [name]_parameters. if empty string is given, properly matching TTrees are searched for in the file.
loadObservables	Flag for whether to load observables for file (true; default) or to let user respecify observables.

Definition at line 166 of file BCIntegrate.cxx.

Member Function Documentation

bool BCIntegrate::CheckMarginalizationAvailability ( BCMarginalizationMethod type )

Check availability of integration routine for marginalization.

Returns: availability of marginalization method

Definition at line 582 of file BCIntegrate.cxx.

std::string BCIntegrate::DumpCubaIntegrationMethod ( BCIntegrate::BCCubaMethod type ) const

Return string with the name for a given Cuba integration type.

Parameters

type	code for the Cuba integration type

Returns: string containing the name of the Cuba integration type

Definition at line 1928 of file BCIntegrate.cxx.

std::string BCIntegrate::DumpCubaIntegrationMethod ( ) const

inline

Return string with the name for the currently set Cuba integration type.

Returns: string containing the name of the Cuba integration type

Definition at line 783 of file BCIntegrate.h.

std::string BCIntegrate::DumpCurrentIntegrationMethod ( ) const

inline

Return string with the name for the currently set integration type.

Returns: string containing the name of the integration type

Definition at line 729 of file BCIntegrate.h.

std::string BCIntegrate::DumpCurrentMarginalizationMethod ( ) const

inline

Return string with the name for the currently set marginalization type.

Returns: string containing the name of the marginalization type

Definition at line 747 of file BCIntegrate.h.

std::string BCIntegrate::DumpCurrentOptimizationMethod ( ) const

inline

Return string with the name for the currently set optimization type.

Returns: string containing the name of the optimization type

Definition at line 765 of file BCIntegrate.h.

std::string BCIntegrate::DumpIntegrationMethod ( BCIntegrate::BCIntegrationMethod type ) const

Return string with the name for a given integration type.

Parameters

type	code for the integration type

Returns: string containing the name of the integration type

Definition at line 1871 of file BCIntegrate.cxx.

std::string BCIntegrate::DumpMarginalizationMethod ( BCIntegrate::BCMarginalizationMethod type ) const

Return string with the name for a given marginalization type.

Parameters

type	code for the marginalization type

Returns: string containing the name of the marginalization type

Definition at line 1890 of file BCIntegrate.cxx.

std::string BCIntegrate::DumpOptimizationMethod ( BCIntegrate::BCOptimizationMethod type ) const

Return string with the name for a given optimization type.

Parameters

type	code for the optimization type

Returns: string containing the name of the optimization type

Definition at line 1909 of file BCIntegrate.cxx.

std::string BCIntegrate::DumpUsedIntegrationMethod ( ) const

inline

Return string with the name for the previously used integration type.

Returns: string containing the name of the integration type

Definition at line 735 of file BCIntegrate.h.

std::string BCIntegrate::DumpUsedMarginalizationMethod ( ) const

inline

Return string with the name for the marginalization type used.

Returns: string containing the name of the marginalization type

Definition at line 753 of file BCIntegrate.h.

std::string BCIntegrate::DumpUsedOptimizationMethod ( ) const

inline

Return string with the name for the optimization type used to find the current mode.

Returns: string containing the name of the optimization type

Definition at line 771 of file BCIntegrate.h.

virtual double BCIntegrate::Eval ( const std::vector< double > & x )

pure virtual

Evaluate the unnormalized probability at a point in parameter space.

Method needs to be overloaded by the user.

Parameters

x	The point in parameter space

Returns: The unnormalized probability

Implemented in BCModel.

std::vector< double > BCIntegrate::FindMode ( std::vector< double > start = std::vector<double>() )

Do the mode finding using a method set via SetOptimizationMethod.

Default is Minuit. The mode can be extracted using the GetBestFitParameters() method.

A starting point for the mode finding can be specified for Minuit. If not specified, the previously found maximum (typically from marginalization) is used as an initial point. If that is not available, then the Minuit default will be used (center of the parameter space).

Returns: The mode found.

Note: The result may not coincide with the result of GetBestFitParameters() if a previous optimization found a better value.

Definition at line 1022 of file BCIntegrate.cxx.

std::vector< double > BCIntegrate::FindMode	(	BCIntegrate::BCOptimizationMethod	optmethod,
		std::vector< double >	start = `std::vector<double>()`
	)

Find mode using a specific method.

The original method will be reset.

Parameters

optmethod	the optimization method
start	the starting point for the optimization algorithm

Returns: the mode

See also: std::vector<double> FindMode(std::vector<double> start = std::vector<double>(0));

Definition at line 1090 of file BCIntegrate.cxx.

double BCIntegrate::GetAbsolutePrecision ( ) const

inline

Returns: Requested absolute precision of the numerical integration

Definition at line 321 of file BCIntegrate.h.

const std::vector< double > & BCIntegrate::GetBestFitParameters ( ) const

virtual

Returns: vector of parameter and observable values at the mode.

Reimplemented from BCEngineMCMC.

Definition at line 274 of file BCIntegrate.cxx.

std::string BCIntegrate::GetBestFitSummary ( unsigned i ) const

protectedvirtual

Get string summarizing best fit for single variable.

Parameters

i	Index of variable to summarize.

Reimplemented from BCEngineMCMC.

Definition at line 2032 of file BCIntegrate.cxx.

const BCCubaOptions::Cuhre& BCIntegrate::GetCubaCuhreOptions ( ) const

inline

Returns: Options used for integration with CUBA's Cuhre

Definition at line 346 of file BCIntegrate.h.

const BCCubaOptions::Divonne& BCIntegrate::GetCubaDivonneOptions ( ) const

inline

Returns: Options used for integration with CUBA's Divonne

Definition at line 341 of file BCIntegrate.h.

BCCubaMethod BCIntegrate::GetCubaIntegrationMethod ( ) const

inline

Returns: Cuba Integration method

Definition at line 326 of file BCIntegrate.h.

const BCCubaOptions::Suave& BCIntegrate::GetCubaSuaveOptions ( ) const

inline

Returns: Options used for integration with CUBA's Suave

Definition at line 336 of file BCIntegrate.h.

const BCCubaOptions::Vegas& BCIntegrate::GetCubaVegasOptions ( ) const

inline

Returns: Options used for integration with CUBA's Vegas

Definition at line 331 of file BCIntegrate.h.

double BCIntegrate::GetError ( ) const

inline

Returns: The uncertainty in the most recent Monte Carlo integration

Definition at line 411 of file BCIntegrate.h.

double BCIntegrate::GetIntegral ( ) const

inline

Returns: The integral.

Definition at line 259 of file BCIntegrate.h.

BCIntegrate::BCIntegrationMethod BCIntegrate::GetIntegrationMethod ( ) const

inline

Returns: The current integration method

Definition at line 269 of file BCIntegrate.h.

double BCIntegrate::GetLogMaximum ( ) const

inlinevirtual

Returns the posterior at the mode.

Returns: the posterior.

Reimplemented from BCEngineMCMC.

Definition at line 442 of file BCIntegrate.h.

BCIntegrate::BCMarginalizationMethod BCIntegrate::GetMarginalizationMethod ( ) const

inline

Returns: The current marginalization method

Definition at line 274 of file BCIntegrate.h.

TMinuitMinimizer& BCIntegrate::GetMinuit ( )

inline

Returns: Minuit minimizer used for mode finding. The object is in a valid but unusable state if no mode finding has been performed.

Definition at line 418 of file BCIntegrate.h.

int BCIntegrate::GetNIterations ( ) const

inline

Returns: The number of iterations for the most recent Monte Carlo integration

Definition at line 311 of file BCIntegrate.h.

int BCIntegrate::GetNIterationsMax ( ) const

inline

Returns: The number of maximum iterations for integration

Definition at line 301 of file BCIntegrate.h.

int BCIntegrate::GetNIterationsMin ( ) const

inline

Returns: The number of minimum iterations for integration

Definition at line 296 of file BCIntegrate.h.

int BCIntegrate::GetNIterationsPrecisionCheck ( ) const

inline

Returns: The interval for checking precision in integration

Definition at line 306 of file BCIntegrate.h.

BCIntegrate::BCOptimizationMethod BCIntegrate::GetOptimizationMethod ( ) const

inline

Returns: The current optimization method

Definition at line 264 of file BCIntegrate.h.

std::vector< double > BCIntegrate::GetProposalPointSA	(	const std::vector< double > &	x,
		int	t
	)		const

Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm.

Delegates the generation to the appropriate method according to fSASchedule.

Parameters

x	last state.
t	time iterator to determine current temperature.

Definition at line 1282 of file BCIntegrate.cxx.

std::vector< double > BCIntegrate::GetProposalPointSABoltzmann	(	const std::vector< double > &	x,
		int	t
	)		const

Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm.

This method is used for Boltzmann annealing schedule.

Parameters

x	last state.
t	time iterator to determine current temperature.

Definition at line 1294 of file BCIntegrate.cxx.

std::vector< double > BCIntegrate::GetProposalPointSACauchy	(	const std::vector< double > &	x,
		int	t
	)		const

Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm.

This method is used for Cauchy annealing schedule.

Parameters

x	last state.
t	time iterator to determine current temperature.

Definition at line 1314 of file BCIntegrate.cxx.

std::vector< double > BCIntegrate::GetProposalPointSACustom	(	const std::vector< double > &	x,
		int	t
	)		const

virtual

Generates a new state in a neighbourhood around x that is to be accepted or rejected by the Simulated Annealing algorithm.

This is a virtual method to be overridden by a user-defined custom proposal function.

Parameters

x	last state.
t	time iterator to determine current temperature.

Definition at line 1356 of file BCIntegrate.cxx.

double BCIntegrate::GetRandomPoint ( std::vector< double > & x )

Fills a vector of (flat) random numbers in the limits of the parameters and returns the probability at that point.

Parameters

x	A vector of doubles to fill

Returns: The (unnormalized) probability at the random point

Definition at line 1006 of file BCIntegrate.cxx.

void BCIntegrate::GetRandomVectorInParameterSpace ( std::vector< double > & x ) const

Fills a vector of random numbers x[i] between fMin[i] and fMax[i] into a vector.

Parameters

x	A vector of doubles to fill

Definition at line 1013 of file BCIntegrate.cxx.

double BCIntegrate::GetRelativePrecision ( ) const

inline

Returns: Requested relative precision of the numerical integation

Definition at line 316 of file BCIntegrate.h.

BCIntegrate::BCSASchedule BCIntegrate::GetSASchedule ( ) const

inline

Returns: The Simulated Annealing schedule

Definition at line 279 of file BCIntegrate.h.

double BCIntegrate::GetSAT0 ( ) const

inline

Returns the Simulated Annealing starting temperature.

Definition at line 423 of file BCIntegrate.h.

double BCIntegrate::GetSATmin ( ) const

inline

Returns the Simulated Annealing threshhold temperature.

Definition at line 428 of file BCIntegrate.h.

TH1 * BCIntegrate::GetSlice	(	std::vector< unsigned >	indices,
		unsigned &	nIterations,
		double &	log_max_val,
		const std::vector< double >	parameters = `std::vector<double>(0)`,
		int	nbins = `0`,
		bool	normalize = `true`
	)

Returns a one-dimensional slice of the pdf at the point and along a specific direction.

Parameters

indices	The indices of the model parameters along which the slice is calculated.
nIterations	Add the number of posterior evaluations performed.
log_max_val	Stores the log of the maximum value before normalizing
parameters	The point at which the other parameters are fixed.
nbins	The number of bins of the 1D-histogram.
normalize	Flag for turning on normalization of histogram.

Returns: The slice histogram.

Definition at line 840 of file BCIntegrate.cxx.

TH1* BCIntegrate::GetSlice	(	const std::string &	name,
		unsigned &	nIterations,
		double &	log_max_val,
		const std::vector< double >	parameters = `std::vector<double>(0)`,
		int	nbins = `0`,
		bool	normalize = `true`
	)

inline

Returns a one-dimensional slice of the pdf at the point and along a specific direction.

Parameters

name	The name of the model parameter along which the slice is calculated.
nIterations	Add the number of posterior evaluations performed.
log_max_val	Stores the log of the maximum value before normalizing
parameters	The point at which the other parameters are fixed.
nbins	The number of bins of the 1D-histogram.
normalize	Flag for turning on normalization of histogram.

Returns: The 1D slice.

Definition at line 369 of file BCIntegrate.h.

TH1* BCIntegrate::GetSlice	(	unsigned	index,
		unsigned &	nIterations,
		double &	log_max_val,
		const std::vector< double >	parameters = `std::vector<double>(0)`,
		int	nbins = `0`,
		bool	normalize = `true`
	)

inline

Returns a one-dimensional slice of the pdf at the point and along a specific direction.

Parameters

index	The index of the model parameter along which the slice is calculated.
nIterations	Add the number of posterior evaluations performed.
log_max_val	Stores the log of the maximum value before normalizing
parameters	The point at which the other parameters are fixed.
nbins	The number of bins of the 1D-histogram.
normalize	Flag for turning on normalization of histogram.

Returns: The 1D slice.

Definition at line 381 of file BCIntegrate.h.

TH2* BCIntegrate::GetSlice	(	const std::string &	name1,
		const std::string &	name2,
		unsigned &	nIterations,
		double &	log_max_val,
		const std::vector< double >	parameters = `std::vector<double>(0)`,
		int	nbins = `0`,
		bool	normalize = `true`
	)

inline

Returns a two-dimensional slice of the pdf at the point and along two specified directions.

Parameters

name1	The name of the first model parameter along which the slice is calculated.
name2	The name of the second model parameter along which the slice is calculated.
nIterations	Add the number of posterior evaluations performed.
log_max_val	Stores the log of the maximum value before normalizing
parameters	The point at which the other parameters are fixed.
nbins	The number of bins on each axis of the 2D-histogram.
normalize	Flag for turning on normalization of histogram.

Returns: The 2D slice.

Definition at line 394 of file BCIntegrate.h.

TH2 * BCIntegrate::GetSlice	(	unsigned	index1,
		unsigned	index2,
		unsigned &	nIterations,
		double &	log_max_val,
		const std::vector< double >	parameters = `std::vector<double>(0)`,
		int	nbins = `0`,
		bool	normalize = `true`
	)

Returns a two-dimensional slice of the pdf at the point and along two specified directions.

Parameters

index1	The index of the first model parameter along which the slice is calculated.
index2	The index of the second model parameter along which the slice is calculated.
nIterations	Add the number of posterior evaluations performed.
log_max_val	Stores the log of the maximum value before normalizing
parameters	The point at which the other parameters are fixed.
nbins	The number of bins on each axis of the 2D-histogram.
normalize	Flag for turning on normalization of histogram.

Returns: The 2D slice.

Definition at line 998 of file BCIntegrate.cxx.

double BCIntegrate::Integrate ( BCIntegrationMethod intmethod )

Does the integration over the un-normalized probability.

Parameters

intmethod The integration method to used

Returns: The normalization value

Definition at line 385 of file BCIntegrate.cxx.

double BCIntegrate::Integrate ( )

Perform the integration.

Returns: the integral

Definition at line 306 of file BCIntegrate.cxx.

double BCIntegrate::Integrate	(	BCIntegrationMethod	type,
		tRandomizer	randomizer,
		tEvaluator	evaluator,
		tIntegralUpdater	updater,
		std::vector< double > &	sums
	)

Does the integration over the un-normalized probability.

Parameters

type	The integration method to used (for printing out status updates by name)
randomizer	Pointer to function to choose next random point
evaluator	Pointer to function to evaluate point
updater	Pointer to function to update integral and precision
sums	Vector of doubles holding values used in integral calculation

Returns: The integral value

Definition at line 490 of file BCIntegrate.cxx.

double BCIntegrate::IntegrateLaplace ( )

Integrate using the Laplace approximation.

Take the result of a previous successful minuit run to estimate the covariance matrix. Else it runs minuit (again).

Returns: the integral on the log(!) scale. Note that is is different from the other integration methods that work on the linear scale.

Definition at line 1782 of file BCIntegrate.cxx.

virtual double BCIntegrate::LogEval ( const std::vector< double > & x )

inlinevirtual

Evaluate the natural logarithm of the Eval function.

For better numerical stability, this method should also be overloaded by the user.

Parameters

x	The point in parameter space

Returns: log(Eval(x))

Implements BCEngineMCMC.

Reimplemented in BCModel.

Definition at line 555 of file BCIntegrate.h.

void BCIntegrate::LogOutputAtEndOfIntegration	(	double	integral,
		double	absprecision,
		double	relprecision,
		int	nIterations
	)

protected

Helper method to output at end of integration.

Definition at line 475 of file BCIntegrate.cxx.

void BCIntegrate::LogOutputAtIntegrationStatusUpdate	(	BCIntegrationMethod	type,
		double	integral,
		double	absprecision,
		int	nIterations
	)

protected

Helper method to output integration status.

Definition at line 484 of file BCIntegrate.cxx.

void BCIntegrate::LogOutputAtStartOfIntegration	(	BCIntegrationMethod	type,
		BCCubaMethod	cubatype
	)

protected

Helper method to output at beginning of integration.

Definition at line 420 of file BCIntegrate.cxx.

int BCIntegrate::MarginalizeAll ( )

Marginalize all probabilities wrt.

single parameters and all combinations of two parameters. The individual distributions can be retrieved using the GetMarginalized method.

Returns: Total number of marginalized distributions

Definition at line 635 of file BCIntegrate.cxx.

int BCIntegrate::MarginalizeAll ( BCIntegrate::BCMarginalizationMethod margmethod )

Marginalize all probabilities wrt.

single parameters and all combinations of two parameters. The individual distributions can be retrieved using the GetMarginalized method.

Parameters

margmethod the marginalization method.

Returns: Total number of marginalized distributions

Definition at line 821 of file BCIntegrate.cxx.

virtual void BCIntegrate::MarginalizePostprocess ( )

inlinevirtual

Method executed after marginalization.

User's code should be provided via overloading in the derived class

Definition at line 616 of file BCIntegrate.h.

virtual void BCIntegrate::MarginalizePreprocess ( )

inlinevirtual

Method executed for before marginalization.

User's code should be provided via overloading in the derived class

Reimplemented in BCFitter.

Definition at line 610 of file BCIntegrate.h.

double BCIntegrate::Normalize ( )

inline

Performs integration.

Definition at line 560 of file BCIntegrate.h.

void BCIntegrate::PrintMarginalizationSummary ( ) const

protectedvirtual

Print marginalization to log.

Reimplemented from BCEngineMCMC.

Definition at line 1997 of file BCIntegrate.cxx.

void BCIntegrate::ResetResults ( )

virtual

Reset all information on the best-fit parameters.

Reimplemented from BCEngineMCMC.

Definition at line 292 of file BCIntegrate.cxx.

double BCIntegrate::SAHelperGetRadialCauchy ( ) const

Generates the radial part of a n-dimensional Cauchy distribution.

Helper function for Cauchy annealing.

Definition at line 1404 of file BCIntegrate.cxx.

std::vector< double > BCIntegrate::SAHelperGetRandomPointOnHypersphere ( ) const

Generates a uniform distributed random point on the surface of a fNvar-dimensional Hypersphere.

Used as a helper to generate proposal points for Cauchy annealing.

Definition at line 1363 of file BCIntegrate.cxx.

double BCIntegrate::SAHelperSinusToNIntegral	(	int	dim,
		double	theta
	)		const

Returns the Integral of sin^dim from 0 to theta.

Helper function needed for generating Cauchy distributions.

Definition at line 1460 of file BCIntegrate.cxx.

double BCIntegrate::SATemperature ( double t ) const

Temperature annealing schedule for use with Simulated Annealing.

Delegates to the appropriate method according to fSASchedule.

Parameters

t	iterator for lowering the temperature over time.

Definition at line 1251 of file BCIntegrate.cxx.

double BCIntegrate::SATemperatureBoltzmann ( double t ) const

Temperature annealing schedule for use with Simulated Annealing.

This method is used for Boltzmann annealing schedule.

Parameters

t	iterator for lowering the temperature over time.

Definition at line 1263 of file BCIntegrate.cxx.

double BCIntegrate::SATemperatureCauchy ( double t ) const

Temperature annealing schedule for use with Simulated Annealing.

This method is used for Cauchy annealing schedule.

Parameters

t	iterator for lowering the temperature over time.

Definition at line 1269 of file BCIntegrate.cxx.

double BCIntegrate::SATemperatureCustom ( double t ) const

virtual

Temperature annealing schedule for use with Simulated Annealing.

This is a virtual method to be overridden by a user-defined custom temperature schedule.

Parameters

t	iterator for lowering the temperature over time.

Definition at line 1275 of file BCIntegrate.cxx.

void BCIntegrate::SetBestFitParameters ( const std::vector< double > & x )

Set best fit parameters values.

Parameters

x	Parameter set to designate as best fit parameters.

Definition at line 414 of file BCIntegrate.cxx.

void BCIntegrate::SetBestFitParameters	(	const std::vector< double > &	x,
		const double &	new_value,
		double &	old_value
	)

Parameters

x	Set best-fit parameters to `x` if `new_value >= old_value`
new_value	is the value of the function at `x`
old_value	is the old best fit value, updated to new_value, if it is larger

Definition at line 405 of file BCIntegrate.cxx.

void BCIntegrate::SetCubaOptions ( const BCCubaOptions::Vegas & options )

inline

Set options for CUBA's Vegas.

Parameters

options Options for CUBA

Definition at line 506 of file BCIntegrate.h.

void BCIntegrate::SetCubaOptions ( const BCCubaOptions::Suave & options )

inline

Set options for CUBA's Suave.

Parameters

options Options for CUBA

Definition at line 512 of file BCIntegrate.h.

void BCIntegrate::SetCubaOptions ( const BCCubaOptions::Divonne & options )

inline

Set options for CUBA's Divonne.

Parameters

options Options for CUBA

Definition at line 518 of file BCIntegrate.h.

void BCIntegrate::SetCubaOptions ( const BCCubaOptions::Cuhre & options )

inline

Set options for CUBA's Cuhre.

Parameters

options Options for CUBA

Definition at line 524 of file BCIntegrate.h.

void BCIntegrate::SetFlagIgnorePrevOptimization ( bool flag )

inline

Parameters

flag	Flag whether or not to ignore result of previous mode finding

Definition at line 451 of file BCIntegrate.h.

void BCIntegrate::SetIntegrationMethod ( BCIntegrate::BCIntegrationMethod method )

Parameters

method The current integration method

Definition at line 1488 of file BCIntegrate.cxx.

void BCIntegrate::SetMarginalizationMethod ( BCIntegrate::BCMarginalizationMethod method )

inline

Parameters

method The current marginalization method

Definition at line 465 of file BCIntegrate.h.

void BCIntegrate::SetNIterationsMax ( int niterations )

inline

Parameters

niterations The maximum number of iterations for integration

Definition at line 480 of file BCIntegrate.h.

void BCIntegrate::SetNIterationsMin ( int niterations )

inline

Parameters

niterations The maximum number of iterations for integration

Definition at line 475 of file BCIntegrate.h.

void BCIntegrate::SetNIterationsPrecisionCheck ( int niterations )

inline

Parameters

niterations interval for checking precision in integration routines

Definition at line 485 of file BCIntegrate.h.

void BCIntegrate::SetOptimizationMethod ( BCIntegrate::BCOptimizationMethod method )

inline

Parameters

method The current optimization method

Definition at line 456 of file BCIntegrate.h.

void BCIntegrate::SetRelativePrecision ( double relprecision )

inline

Parameters

relprecision The relative precision envisioned for Monte Carlo integration

Definition at line 491 of file BCIntegrate.h.

void BCIntegrate::SetSASchedule ( BCIntegrate::BCSASchedule schedule )

inline

Parameters

schedule The Simulated Annealing schedule

Definition at line 470 of file BCIntegrate.h.

void BCIntegrate::SetSAT0 ( double T0 )

inline

Set starting temperature for Simulated Annealing.

Parameters

T0	starting temperature.

Definition at line 530 of file BCIntegrate.h.

void BCIntegrate::SetSATmin ( double Tmin )

inline

Set threshold temperature for Simulated Annealing.

Parameters

Tmin	threshold temperature.

Definition at line 536 of file BCIntegrate.h.

Member Data Documentation

double BCIntegrate::fSALogProb

protected

Log probability of current simulated annealing iteration.

Definition at line 858 of file BCIntegrate.h.

int BCIntegrate::fSANIterations

protected

Number of iterations for simualted annealing.

Definition at line 850 of file BCIntegrate.h.

double BCIntegrate::fSATemperature

protected

Current temperature of simulated annealing algorithm.

Definition at line 854 of file BCIntegrate.h.

std::vector<double> BCIntegrate::fSAx

protected

Current simulated annealing parameter point.

Definition at line 862 of file BCIntegrate.h.

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

/root/bat/BAT/BCIntegrate.h
/root/bat/src/BCIntegrate.cxx

Public Types

Public Member Functions

Protected Member Functions

Protected Attributes

Member functions (miscellaneous methods)

Additional Inherited Members

Detailed Description

Member Enumeration Documentation

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation