Euclid::MathUtils::InverseCumulative< TKnot, E > Class Template Reference

#include <InverseCumulative.h>

Public Member Functions
	InverseCumulative (std::vector< TKnot > knots, std::vector< double > pdf)
TKnot	operator() (double p) const

Detailed Description

template<typename TKnot, typename E = void>
class Euclid::MathUtils::InverseCumulative< TKnot, E >

Model an inverse cumulative, used to sample a value given a PDF

Template Parameters

TKnot	Knot type, may be discrete or continuous
E	Used by the discrete and continuous specializations.

The PDF is linearly interpolated. This implies that the CDF is computed deriving analytically each segment into a quadratic function.

Definition at line 39 of file InverseCumulative.h.

Constructor & Destructor Documentation

template<typename TKnot , typename E = void>

Euclid::MathUtils::InverseCumulative< TKnot, E >::InverseCumulative	(	std::vector< TKnot >	knots,
		std::vector< double >	pdf
	)

Constructor

Parameters

knots	PDF knots. For continuous distributions, they must be in order. Discrete distributions do not need to be sorted.
pdf	Distribution PDF

Member Function Documentation

template<typename TKnot , typename E = void>

TKnot Euclid::MathUtils::InverseCumulative< TKnot, E >::operator()

(

double

p

)

const

Return the first value where the CDF has a value >= p For continuous distributions each segment of the PDF is assumed to be linearly interpolated, so the integral of the PDF is used to interpolate the knot value.

---

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

• MathUtils/MathUtils/helpers/InverseCumulative.h