#include <CensoredQuantileRegression.hh>

Inheritance diagram for npstat::CensoredQuantileRegressionOnHisto< Numeric, Num2 >:

Public Member Functions
	CensoredQuantileRegressionOnHisto (const HistoND< Numeric > &histo, const ArrayND< Num2 > &efficiency, const ArrayND< Num2 > &cutoff, const QuadraticOrthoPolyND &poly, double quantile, double valueLimit)

virtual void	process (const unsigned *index, unsigned indexLen, unsigned long linearIndex, const Numeric &value)

Public Member Functions inherited from npstat::QuantileRegressionOnHisto< Numeric >
	QuantileRegressionOnHisto (const HistoND< Numeric > &histo, const QuadraticOrthoPolyND &poly, double quantile)

const HistoND< Numeric > &	getHisto () const

virtual void	clear ()

virtual double	result ()

Public Member Functions inherited from npstat::QuantileRegressionBase< double >
	QuantileRegressionBase (const QuadraticOrthoPolyND &poly, double quantile)

unsigned	dim () const

double	quantile () const

const QuadraticOrthoPolyND &	getPoly () const

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

void	setRegressionBox (const BoxND< double > &box)

void	setLinearMapping (const double location, const double scale, unsigned locationAndScaleArraySize)

void	empiricalC0 (double c0, double c0Uncertainty, double *npoints)

double	linearLoss (const double *coeffs, unsigned nCoeffs)

Additional Inherited Members
Protected Member Functions inherited from npstat::QuantileRegressionBase< double >
void	resetAccumulators ()

Protected Attributes inherited from npstat::QuantileRegressionOnHisto< Numeric >
const HistoAxis &	lastAxis_

const double	halfbin_

const Numeric	zero_

const HistoND< Numeric > &	histo_

BoxND< int >	binBox_

Protected Attributes inherited from npstat::QuantileRegressionBase< double >
const QuadraticOrthoPolyND &	poly_

BoxND< double >	regressionBox_

std::vector< double >	location_

std::vector< double >	scale_

std::vector< double >	coords_

std::vector< double >	coeffs_

long double	loss_

const double	quantile_

const double	onemq_

const unsigned	mydim_

Detailed Description

template<typename Numeric, typename Num2>
class npstat::CensoredQuantileRegressionOnHisto< Numeric, Num2 >

This class assumes that the histogram is filled with actual event counts and that the quantity which is being regressed is used for the last histogram axis. The dimensionality of the histogram therefore must be larger by 1 than the dimensionality of the QuadraticOrthoPolyND object. Histogram overflow bins will be ignored.

Constructor & Destructor Documentation

◆ CensoredQuantileRegressionOnHisto()

template<typename Numeric , typename Num2 >

npstat::CensoredQuantileRegressionOnHisto< Numeric, Num2 >::CensoredQuantileRegressionOnHisto	(	const HistoND< Numeric > &	histo,
		const ArrayND< Num2 > &	efficiency,
		const ArrayND< Num2 > &	cutoff,
		const QuadraticOrthoPolyND &	poly,
		double	quantile,
		double	valueLimit
	)

It is assumed that the arrays "efficiency" and "cutoff" contain the efficiency and cutoff values for the predictor values placed in the center of the argument histogram bins (all histogram dimensions except last are predictor dimensions). Therefore, they should be bin-compatible with the histogram binning constructed without last axis.

"valueLimit" has the same meaning as in the class CensoredQuantileRegressionOnKDTree. Naturally, this value should be outside (or exactly on the boundary) of the range covered by the last histogram axis.

Member Function Documentation

◆ process()

template<typename Numeric , typename Num2 >

virtual void npstat::CensoredQuantileRegressionOnHisto< Numeric, Num2 >::process	(	const unsigned *	index,
		unsigned	indexLen,
		unsigned long	linearIndex,
		const Numeric &	value
	)

virtual

Method from AbsArrayProjector that we have to implement

Reimplemented from npstat::QuantileRegressionOnHisto< Numeric >.

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

stat/CensoredQuantileRegression.hh

Public Member Functions