npsi::JohnsonFit Class Reference

#include <minuitFitJohnsonCurves.hh>

Public Member Functions
template<class Numeric , class Axis >
	JohnsonFit (const npstat::HistoND< Numeric, Axis > &histo, double qmin, double qmax, double minlog)
bool	converged () const
double	mean () const
double	sigma () const
double	skewness () const
double	kurtosis () const
unsigned	minFitBin () const
unsigned	maxFitBin () const

Detailed Description

Fitting Johnson system of curves to histogrammed data using the maximum likelihood method

Constructor & Destructor Documentation

JohnsonFit()

template<class Numeric , class Axis >

npsi::JohnsonFit::JohnsonFit	(	const npstat::HistoND< Numeric, Axis > &	histo,
		double	qmin,
		double	qmax,
		double	minlog
	)

In the constructor, arguments are as follows:

"histo" is the 1-d histogram to fit.

"qmin" and "qmax" are the quantiles which define the fitted region. For example, set them to 0.05 and 0.95, respectively, if you would like to fit central 90% of the distribution.

"minlog" limits the contribution of any single bin into the log-likelihood from below. No bin can contribute less then "minlog" even if the value of the fitted density for that bin is 0. Typically, this parameter should be set to a negative number with a reasonably large magnitude, on the order of log(DBL_MIN).

Member Function Documentation

converged()

bool npsi::JohnsonFit::converged ( ) const

inline

Check fit convergence. If the fit did not converge, the results will be set by simple array-based calculations.

mean()

double npsi::JohnsonFit::mean ( ) const

inline

Examine fitted value

minFitBin()

unsigned npsi::JohnsonFit::minFitBin ( ) const

inline

Figure out the region fitted. The minimum bin is included and the maximum bin is excluded from the fit.

---

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

• interfaces/minuitFitJohnsonCurves.hh