The class that represents a distribution. More...

#include <Distribution.h>

Collaboration diagram for ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >:

Public Member Functions
KOKKOS_INLINE_FUNCTION	Distribution (const T *par_p)
	Constructor for the Distribution class. More...

KOKKOS_INLINE_FUNCTION	~Distribution ()
	Destructor for the Distribution class. More...

KOKKOS_INLINE_FUNCTION T	getPdf (T x, unsigned int d) const
	A wrapper to change the signature arguments of pdf in each dimension d from (x, d, par) to (x, d). More...

KOKKOS_INLINE_FUNCTION T	getCdf (T x, unsigned int d) const
	A wrapper to change the signature arguments of cdf in each dimension d from (x, d, par) to (x, d). More...

KOKKOS_INLINE_FUNCTION T	getEstimate (T x, unsigned int d) const
	A wrapper to change the signature arguments of estimate in each dimension d from (x, d, par) to (x, d). More...

KOKKOS_INLINE_FUNCTION T	getObjFunc (T x, unsigned int d, T u) const

KOKKOS_INLINE_FUNCTION T	getDerObjFunc (T x, unsigned int d) const

KOKKOS_INLINE_FUNCTION T	getFullPdf (ippl::Vector< T, Dim > x) const

Public Attributes
T	par_m [DimP]

DistributionFunctions::PDF	pdf_m

DistributionFunctions::CDF	cdf_m

DistributionFunctions::Estimate	estimate_m

Detailed Description

template<typename T, unsigned Dim, unsigned DimP, typename DistributionFunctions>
class ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >

The class that represents a distribution.

Template Parameters

T	Datatype.
Dim	Dimensionality of sample space.
DimP	Dimensionality of the parameter array.
PDF	Struct type for the PDF (Probability Density Function).
CDF	Struct type for the CDF (Cumulative Distribution Function).
ESTIMATE	Struct type for the ESTIMATE function.

Definition at line 33 of file ippl/src/Random/Distribution.h.

Constructor & Destructor Documentation

◆ Distribution()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::Distribution ( const T * par_p )

inline

Constructor for the Distribution class.

Parameters

par_	Pointer to the parameter array.

Definition at line 51 of file ippl/src/Random/Distribution.h.

References ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::par_m.

◆ ~Distribution()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::~Distribution ( )

inline

Destructor for the Distribution class.

Definition at line 61 of file ippl/src/Random/Distribution.h.

Member Function Documentation

◆ getCdf()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION T ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getCdf	(	T	x,
		unsigned int	d
	)		const

inline

A wrapper to change the signature arguments of cdf in each dimension d from (x, d, par) to (x, d).

Definition at line 75 of file ippl/src/Random/Distribution.h.

References ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::cdf_m, and ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::par_m.

Referenced by ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getObjFunc(), and ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::updateBounds().

◆ getDerObjFunc()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION T ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getDerObjFunc	(	T	x,
		unsigned int	d
	)		const

inline

Returns: Derivative of the objective function that is used in inverse transform sampling, i.e. d(obj)/dx = pdf(x)

Definition at line 100 of file ippl/src/Random/Distribution.h.

References ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getPdf().

Referenced by ippl::random::detail::NewtonRaphson< T, Distribution >::solve().

Here is the call graph for this function:

◆ getEstimate()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION T ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getEstimate	(	T	x,
		unsigned int	d
	)		const

inline

A wrapper to change the signature arguments of estimate in each dimension d from (x, d, par) to (x, d).

Definition at line 83 of file ippl/src/Random/Distribution.h.

References ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::estimate_m, and ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::par_m.

Referenced by ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::fill_random< GeneratorPool >::operator()().

◆ getFullPdf()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION T ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getFullPdf ( ippl::Vector< T, Dim > x ) const

inline

Returns: Total pdf given uncorrelated pdf in each dimension. i.e. total_pdf = pdf(x_1) * pdf(x_2) * ... pdf(x_N).

Definition at line 108 of file ippl/src/Random/Distribution.h.

References Dim, and ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getPdf().

Here is the call graph for this function:

◆ getObjFunc()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION T ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getObjFunc	(	T	x,
		unsigned int	d,
		T	u
	)		const

inline

Returns: Objective function that is used in inverse transform sampling, i.e. obj = cdf(x)-u. Here u is uniformly distributed on [0, 1] and x is the sample of target distribution.

Definition at line 92 of file ippl/src/Random/Distribution.h.

References ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getCdf().

Referenced by ippl::random::detail::NewtonRaphson< T, Distribution >::solve().

Here is the call graph for this function:

◆ getPdf()

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

KOKKOS_INLINE_FUNCTION T ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getPdf	(	T	x,
		unsigned int	d
	)		const

inline

A wrapper to change the signature arguments of pdf in each dimension d from (x, d, par) to (x, d).

Definition at line 67 of file ippl/src/Random/Distribution.h.

References ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::par_m, and ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::pdf_m.

Referenced by ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getDerObjFunc(), and ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getFullPdf().

Member Data Documentation

◆ cdf_m

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

DistributionFunctions::CDF ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::cdf_m

Definition at line 44 of file ippl/src/Random/Distribution.h.

Referenced by ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getCdf().

◆ estimate_m

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

DistributionFunctions::Estimate ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::estimate_m

Definition at line 45 of file ippl/src/Random/Distribution.h.

Referenced by ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getEstimate().

◆ par_m

template<typename T , unsigned Dim, unsigned DimP, typename DistributionFunctions >

T ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::par_m[DimP]

Parameters

par_m	An array of distribution parameters.
pdf_m	PDF of the distribution class as a member functor.
cdf_m	CDF of the distribution class as a member functor.
estimate_m	Estimate of the initial guess for the sampling method as a member functor.

Definition at line 42 of file ippl/src/Random/Distribution.h.

Referenced by ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::Distribution(), ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getCdf(), ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getEstimate(), and ippl::random::Distribution< T, Dim, DimP, DistributionFunctions >::getPdf().