Manages a collection of concentration measurements with statistical analysis. More...

#include <concentrationset.h>

Inheritance diagram for ConcentrationSet:

Collaboration diagram for ConcentrationSet:

Public Member Functions
	ConcentrationSet ()

	ConcentrationSet (const ConcentrationSet &other)

	ConcentrationSet (int n)

ConcentrationSet &	operator= (const ConcentrationSet &other)

void	AppendValue (double value)
	Append a single concentration value.

void	AppendSet (const ConcentrationSet &other)
	Append all values from another set.

void	AppendSet (ConcentrationSet *other)

double	CalculateMean () const
	Calculate arithmetic mean.

double	CalculateStdDev (double mean_value=-999) const
	Calculate standard deviation.

double	CalculateMeanLog () const
	Calculate mean of log-transformed values.

double	CalculateStdDevLog (double mean_value=-999) const
	Calculate standard deviation of log-transformed values.

double	CalculateSSE (double mean_value=-999) const
	Calculate sum of squared errors from mean.

double	CalculateSSELog (double mean_value=-999) const
	Calculate sum of squared errors from log mean.

double	CalculateNorm (double power=2.0) const
	Calculate sum of values raised to a power.

double	CalculateNormLog (double power=2.0) const
	Calculate sum of log-transformed values raised to a power.

double	GetMinimum () const
	Get minimum value.

double	GetMaximum () const
	Get maximum value.

vector< double >	EstimateDistributionParameters (distribution_type dist_type=distribution_type::none)
	Estimate distribution parameters from data.

double	CalculateLogLikelihood (const vector< double > &params=vector< double >(), distribution_type dist_type=distribution_type::none) const
	Calculate log-likelihood of data given distribution parameters.

distribution_type	SelectBestDistribution ()
	Select best-fitting distribution (normal vs lognormal)

Distribution *	GetFittedDistribution ()
	Get fitted distribution (from data)

const Distribution *	GetFittedDistribution () const

Distribution *	GetEstimatedDistribution ()
	Get estimated distribution (for MCMC optimization)

const Distribution *	GetEstimatedDistribution () const

double	GetEstimatedMu () const
	Get/set estimated mu parameter.

void	SetEstimatedMu (double value)

double	GetEstimatedSigma () const
	Get/set estimated sigma parameter.

void	SetEstimatedSigma (double value)

CMBTimeSeries	CreateDataCDF () const
	Create empirical CDF from data.

CMBTimeSeriesSet	CreateCDFComparison (distribution_type dist_type) const
	Create comparison of empirical vs fitted CDF.

CMBTimeSeriesSet	CreateFittedDistribution (distribution_type dist_type) const
	Create fitted distribution visualization.

double	CalculateKolmogorovSmirnovStatistic (distribution_type dist_type) const
	Calculate Kolmogorov-Smirnov statistic.

ConcentrationSet	CreateLogTransformed () const
	Create log-transformed copy.

double	CalculateBoxCoxLogLikelihood (double lambda) const
	Calculate Box-Cox log-likelihood for given lambda.

ConcentrationSet	ApplyBoxCoxTransform (double lambda, bool normalize) const
	Apply Box-Cox transformation.

double	FindOptimalBoxCoxParameter (double min_lambda, double max_lambda, int n_intervals) const
	Find optimal Box-Cox lambda parameter.

vector< unsigned int >	CalculateRanks () const
	Calculate ranks of values (1 = smallest)

Private Attributes
Distribution	fitted_distribution_
	Distribution fitted to observed data.

Distribution	estimated_distribution_
	Distribution being optimized (e.g., in MCMC)

Detailed Description

Manages a collection of concentration measurements with statistical analysis.

Inherits from vector<double> to store concentration values and provides:

Statistical measures (mean, standard deviation, norms)
Distribution fitting (normal, lognormal)
Box-Cox transformations
Goodness-of-fit testing (Kolmogorov-Smirnov)

Definition at line 22 of file concentrationset.h.

Constructor & Destructor Documentation

◆ ConcentrationSet() [1/3]

ConcentrationSet::ConcentrationSet ( )

Definition at line 9 of file concentrationset.cpp.

References vector.

◆ ConcentrationSet() [2/3]

ConcentrationSet::ConcentrationSet ( const ConcentrationSet & other )

Definition at line 16 of file concentrationset.cpp.

References vector.

◆ ConcentrationSet() [3/3]

ConcentrationSet::ConcentrationSet ( int n )

Definition at line 23 of file concentrationset.cpp.

References vector.

Member Function Documentation

◆ AppendSet() [1/2]

void ConcentrationSet::AppendSet ( ConcentrationSet * other )

Definition at line 55 of file concentrationset.cpp.

◆ AppendSet() [2/2]

void ConcentrationSet::AppendSet ( const ConcentrationSet & other )

Append all values from another set.

Definition at line 50 of file concentrationset.cpp.

Referenced by SourceSinkData::DifferentiationPower_Percentage().

◆ AppendValue()

void ConcentrationSet::AppendValue ( double value )

Append a single concentration value.

Definition at line 45 of file concentrationset.cpp.

Referenced by SourceSinkData::ExtractConcentrationSet().

◆ ApplyBoxCoxTransform()

ConcentrationSet ConcentrationSet::ApplyBoxCoxTransform	(	double	lambda,
		bool	normalize
	)		const

Apply Box-Cox transformation.

Parameters

lambda	Transformation parameter
normalize	If true, scale by standard deviation

Definition at line 389 of file concentrationset.cpp.

References CalculateStdDev(), GetMinimum(), and log.

Referenced by Elemental_Profile_Set::ApplyBoxCoxTransform(), CalculateBoxCoxLogLikelihood(), Elemental_Profile_Set::DetectOutliers(), and FindOptimalBoxCoxParameter().

◆ CalculateBoxCoxLogLikelihood()

double ConcentrationSet::CalculateBoxCoxLogLikelihood ( double lambda ) const

Calculate Box-Cox log-likelihood for given lambda.

Definition at line 370 of file concentrationset.cpp.

References ApplyBoxCoxTransform(), CalculateMean(), CalculateStdDev(), and log.

◆ CalculateKolmogorovSmirnovStatistic()

double ConcentrationSet::CalculateKolmogorovSmirnovStatistic ( distribution_type dist_type ) const

Calculate Kolmogorov-Smirnov statistic.

Definition at line 356 of file concentrationset.cpp.

References CreateCDFComparison().

Referenced by FindOptimalBoxCoxParameter().

◆ CalculateLogLikelihood()

double ConcentrationSet::CalculateLogLikelihood	(	const vector< double > &	params = `vector<double>()`,
		distribution_type	dist_type = `distribution_type::none`
	)		const

Calculate log-likelihood of data given distribution parameters.

Parameters

params	Distribution parameters (empty to use fitted_distribution_)
dist_type	Distribution type (none to use fitted_distribution_)

Returns: Log-likelihood value

Definition at line 233 of file concentrationset.cpp.

References Distribution::distribution, Distribution::Eval(), fitted_distribution_, log, none, and Distribution::parameters.

◆ CalculateMean()

double ConcentrationSet::CalculateMean ( ) const

Calculate arithmetic mean.

Definition at line 64 of file concentrationset.cpp.

Referenced by SourceSinkData::ANOVA(), CalculateBoxCoxLogLikelihood(), CalculateSSE(), CalculateStdDev(), CreateCDFComparison(), CreateFittedDistribution(), Elemental_Profile_Set::DetectOutliers(), SourceSinkData::DifferentiationPower(), EstimateDistributionParameters(), SourceSinkData::GrandMean(), and SourceSinkData::t_TestPValue().

◆ CalculateMeanLog()

double ConcentrationSet::CalculateMeanLog ( ) const

Calculate mean of log-transformed values.

Definition at line 142 of file concentrationset.cpp.

References log.

Referenced by SourceSinkData::ANOVA(), CalculateSSELog(), CalculateStdDevLog(), CreateCDFComparison(), CreateFittedDistribution(), SourceSinkData::DifferentiationPower(), EstimateDistributionParameters(), SourceSinkData::GrandMean(), and SourceSinkData::t_TestPValue().

◆ CalculateNorm()

double ConcentrationSet::CalculateNorm ( double power = 2.0 ) const

Calculate sum of values raised to a power.

Definition at line 156 of file concentrationset.cpp.

References power.

◆ CalculateNormLog()

double ConcentrationSet::CalculateNormLog ( double power = 2.0 ) const

Calculate sum of log-transformed values raised to a power.

Definition at line 166 of file concentrationset.cpp.

References log, and power.

◆ CalculateRanks()

vector< unsigned int > ConcentrationSet::CalculateRanks ( ) const

Calculate ranks of values (1 = smallest)

Definition at line 478 of file concentrationset.cpp.

Referenced by SourceSinkData::DifferentiationPower_Percentage().

◆ CalculateSSE()

double ConcentrationSet::CalculateSSE ( double mean_value = -999 ) const

Calculate sum of squared errors from mean.

Definition at line 114 of file concentrationset.cpp.

References CalculateMean().

Referenced by SourceSinkData::ANOVA().

◆ CalculateSSELog()

double ConcentrationSet::CalculateSSELog ( double mean_value = -999 ) const

Calculate sum of squared errors from log mean.

Definition at line 128 of file concentrationset.cpp.

References CalculateMeanLog(), and log.

Referenced by SourceSinkData::ANOVA().

◆ CalculateStdDev()

double ConcentrationSet::CalculateStdDev ( double mean_value = -999 ) const

Calculate standard deviation.

Parameters

mean_value Pre-calculated mean, or -999 to calculate automatically

Definition at line 78 of file concentrationset.cpp.

References CalculateMean().

Referenced by ApplyBoxCoxTransform(), CalculateBoxCoxLogLikelihood(), CreateCDFComparison(), CreateFittedDistribution(), Elemental_Profile_Set::DetectOutliers(), SourceSinkData::DifferentiationPower(), EstimateDistributionParameters(), and SourceSinkData::t_TestPValue().

◆ CalculateStdDevLog()

double ConcentrationSet::CalculateStdDevLog ( double mean_value = -999 ) const

Calculate standard deviation of log-transformed values.

Parameters

mean_value Pre-calculated log mean, or -999 to calculate automatically

Definition at line 96 of file concentrationset.cpp.

References CalculateMeanLog(), and log.

Referenced by SourceSinkData::ANOVA(), CreateCDFComparison(), CreateFittedDistribution(), SourceSinkData::DifferentiationPower(), EstimateDistributionParameters(), and SourceSinkData::t_TestPValue().

◆ CreateCDFComparison()

CMBTimeSeriesSet ConcentrationSet::CreateCDFComparison ( distribution_type dist_type ) const

Create comparison of empirical vs fitted CDF.

Definition at line 287 of file concentrationset.cpp.

References CalculateMean(), CalculateMeanLog(), CalculateStdDev(), CalculateStdDevLog(), CreateDataCDF(), lognormal, normal, and x.

Referenced by CalculateKolmogorovSmirnovStatistic(), Conductor::ExecuteDistributionFitting(), and Conductor::ExecuteKolmogorovSmirnovIndividual().

◆ CreateDataCDF()

CMBTimeSeries ConcentrationSet::CreateDataCDF ( ) const

Create empirical CDF from data.

Definition at line 272 of file concentrationset.cpp.

Referenced by CreateCDFComparison(), and CreateFittedDistribution().

◆ CreateFittedDistribution()

CMBTimeSeriesSet ConcentrationSet::CreateFittedDistribution ( distribution_type dist_type ) const

Create fitted distribution visualization.

Definition at line 321 of file concentrationset.cpp.

References CalculateMean(), CalculateMeanLog(), CalculateStdDev(), CalculateStdDevLog(), CreateDataCDF(), Distribution::Eval(), lognormal, normal, and x.

Referenced by Conductor::ExecuteDistributionFitting().

◆ CreateLogTransformed()

ConcentrationSet ConcentrationSet::CreateLogTransformed ( ) const

Create log-transformed copy.

Definition at line 176 of file concentrationset.cpp.

References log.

◆ EstimateDistributionParameters()

vector< double > ConcentrationSet::EstimateDistributionParameters ( distribution_type dist_type = distribution_type::none )

Estimate distribution parameters from data.

Parameters

dist_type Distribution type (normal or lognormal)

Returns: Vector of parameters [mean, std_dev] or [mu, sigma]

Definition at line 199 of file concentrationset.cpp.

References CalculateMean(), CalculateMeanLog(), CalculateStdDev(), CalculateStdDevLog(), Distribution::distribution, estimated_distribution_, fitted_distribution_, lognormal, none, normal, Distribution::parameters, Distribution::SetDataMean(), Distribution::SetDataSTDev(), and Distribution::SetType().

Referenced by SourceSinkData::AssignAllDistributions().

◆ FindOptimalBoxCoxParameter()

double ConcentrationSet::FindOptimalBoxCoxParameter	(	double	min_lambda,
		double	max_lambda,
		int	n_intervals
	)		const

Find optimal Box-Cox lambda parameter.

Parameters

min_lambda	Lower bound for search
max_lambda	Upper bound for search
n_intervals	Number of intervals for grid search

Returns: Optimal lambda value

Definition at line 430 of file concentrationset.cpp.

References ApplyBoxCoxTransform(), CalculateKolmogorovSmirnovStatistic(), FindOptimalBoxCoxParameter(), GetMaximum(), GetMinimum(), and normal.

Referenced by FindOptimalBoxCoxParameter().

◆ GetEstimatedDistribution() [1/2]

Distribution * ConcentrationSet::GetEstimatedDistribution ( )

inline

Get estimated distribution (for MCMC optimization)

Definition at line 136 of file concentrationset.h.

References estimated_distribution_.

Referenced by SourceSinkData::LogLikelihoodSourceElementalDistributions().

◆ GetEstimatedDistribution() [2/2]

const Distribution * ConcentrationSet::GetEstimatedDistribution ( ) const

inline

Definition at line 137 of file concentrationset.h.

References estimated_distribution_.

◆ GetEstimatedMu()

double ConcentrationSet::GetEstimatedMu ( ) const

inline

Get/set estimated mu parameter.

Definition at line 142 of file concentrationset.h.

References estimated_distribution_, and Distribution::parameters.

◆ GetEstimatedSigma()

double ConcentrationSet::GetEstimatedSigma ( ) const

inline

Get/set estimated sigma parameter.

Definition at line 151 of file concentrationset.h.

References estimated_distribution_, and Distribution::parameters.

◆ GetFittedDistribution() [1/2]

Distribution * ConcentrationSet::GetFittedDistribution ( )

inline

Get fitted distribution (from data)

Definition at line 130 of file concentrationset.h.

References fitted_distribution_.

Referenced by SourceSinkData::AssignAllDistributions().

◆ GetFittedDistribution() [2/2]

const Distribution * ConcentrationSet::GetFittedDistribution ( ) const

inline

Definition at line 131 of file concentrationset.h.

References fitted_distribution_.

◆ GetMaximum()

double ConcentrationSet::GetMaximum ( ) const

Get maximum value.

Definition at line 192 of file concentrationset.cpp.

Referenced by SourceSinkData::BracketTest(), and FindOptimalBoxCoxParameter().

◆ GetMinimum()

double ConcentrationSet::GetMinimum ( ) const

Get minimum value.

Definition at line 187 of file concentrationset.cpp.

Referenced by ApplyBoxCoxTransform(), Elemental_Profile_Set::CheckForNegativeValues(), FindOptimalBoxCoxParameter(), and SelectBestDistribution().

◆ operator=()

ConcentrationSet & ConcentrationSet::operator= ( const ConcentrationSet & other )

Definition at line 30 of file concentrationset.cpp.

References estimated_distribution_, and fitted_distribution_.

◆ SelectBestDistribution()

distribution_type ConcentrationSet::SelectBestDistribution ( )

Select best-fitting distribution (normal vs lognormal)

Returns: Distribution type with higher log-likelihood

Definition at line 259 of file concentrationset.cpp.

References GetMinimum(), lognormal, and normal.

◆ SetEstimatedMu()

void ConcentrationSet::SetEstimatedMu ( double value )

inline

Definition at line 143 of file concentrationset.h.

References estimated_distribution_, and Distribution::parameters.

Referenced by SourceSinkData::SetParameterValue().

◆ SetEstimatedSigma()

void ConcentrationSet::SetEstimatedSigma ( double value )

inline

Definition at line 152 of file concentrationset.h.

References estimated_distribution_, and Distribution::parameters.

Referenced by SourceSinkData::SetParameterValue().

Member Data Documentation

◆ estimated_distribution_

Distribution ConcentrationSet::estimated_distribution_

private

Distribution being optimized (e.g., in MCMC)

Definition at line 220 of file concentrationset.h.

Referenced by EstimateDistributionParameters(), GetEstimatedDistribution(), GetEstimatedDistribution(), GetEstimatedMu(), GetEstimatedSigma(), operator=(), SetEstimatedMu(), and SetEstimatedSigma().

◆ fitted_distribution_

Distribution ConcentrationSet::fitted_distribution_

private

Distribution fitted to observed data.

Definition at line 219 of file concentrationset.h.

Referenced by CalculateLogLikelihood(), EstimateDistributionParameters(), GetFittedDistribution(), GetFittedDistribution(), and operator=().

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

Public Member Functions

Private Attributes

Detailed Description

Constructor & Destructor Documentation

◆ ConcentrationSet() [1/3]

◆ ConcentrationSet() [2/3]

◆ ConcentrationSet() [3/3]

Member Function Documentation

◆ AppendSet() [1/2]

◆ AppendSet() [2/2]

◆ AppendValue()

◆ ApplyBoxCoxTransform()

◆ CalculateBoxCoxLogLikelihood()

◆ CalculateKolmogorovSmirnovStatistic()

◆ CalculateLogLikelihood()

◆ CalculateMean()

◆ CalculateMeanLog()

◆ CalculateNorm()

◆ CalculateNormLog()

◆ CalculateRanks()

◆ CalculateSSE()

◆ CalculateSSELog()

◆ CalculateStdDev()

◆ CalculateStdDevLog()

◆ CreateCDFComparison()

◆ CreateDataCDF()

◆ CreateFittedDistribution()

◆ CreateLogTransformed()

◆ EstimateDistributionParameters()

◆ FindOptimalBoxCoxParameter()

◆ GetEstimatedDistribution() [1/2]

◆ GetEstimatedDistribution() [2/2]

◆ GetEstimatedMu()

◆ GetEstimatedSigma()

◆ GetFittedDistribution() [1/2]

◆ GetFittedDistribution() [2/2]

◆ GetMaximum()

◆ GetMinimum()

◆ operator=()

◆ SelectBestDistribution()

◆ SetEstimatedMu()

◆ SetEstimatedSigma()

Member Data Documentation

◆ estimated_distribution_

◆ fitted_distribution_