A class for inverse transform sampling. More...

#include <InverseTransformSampling.h>

Collaboration diagram for ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >:

Classes
struct	fill_random
	Functor that is used for generating samples. More...

Public Types
using	view_type = typename ippl::detail::ViewType<Vector<T, Dim>, 1>::view_type
using	size_type = ippl::detail::size_type

Public Member Functions
template<class RegionLayout>
	InverseTransformSampling (Distribution &dist_r, Vector< T, Dim > &rmax_r, Vector< T, Dim > &rmin_r, RegionLayout &rlayout_r, size_type &ntotal_r)
	Constructor for InverseTransformSampling class with domain decomposition.
	InverseTransformSampling (Distribution &dist_r, Vector< T, Dim > &rmax_r, Vector< T, Dim > &rmin_r, Vector< T, Dim > &locrmax_r, Vector< T, Dim > &locrmin_r, size_type &ntotal_r)
	Constructor for InverseTransformSampling class without applying domain decomposition..
	InverseTransformSampling (Distribution &dist_r, Vector< T, Dim > &rmax_r, Vector< T, Dim > &rmin_r, size_type &ntotal_r)
	Constructor for InverseTransformSampling class. In this method, we do not consider any domain decomposition.
void	updateBounds (Vector< T, Dim > &rmax, Vector< T, Dim > &rmin, Vector< T, Dim > &locrmax, Vector< T, Dim > &locrmin)
	Updates the sampling bounds and reinitializes internal variables.
void	updateBounds (Vector< T, Dim > &rmax, Vector< T, Dim > &rmin)
	Updates the sampling bounds using the CDF without any domain decomposition.
	~InverseTransformSampling ()
	Deconstructor for InverseTransformSampling class.
KOKKOS_INLINE_FUNCTION size_type	getLocalSamplesNum () const
	Get the local number of samples.
KOKKOS_INLINE_FUNCTION void	setLocalSamplesNum (size_type nlocal)
	Set the local number of particles.
void	generate (view_type view, Kokkos::Random_XorShift64_Pool<> rand_pool64)
	Generate random samples using inverse transform sampling.
void	generate (view_type view, size_type startIndex, size_type endIndex, Kokkos::Random_XorShift64_Pool<> rand_pool64)
	Generate random samples using inverse transform sampling for a specific range of particles.

Public Attributes
Distribution	dist_m
size_type	ntotal_m
Vector< T, Dim >	umin_m
Vector< T, Dim >	umax_m

Private Attributes
size_type	nlocal_m

Detailed Description

template<typename T, unsigned Dim, class DeviceType, class Distribution>
class ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >

A class for inverse transform sampling.

This class performs inverse transform sampling for a given distribution.

Template Parameters

T	Data type.
Dim	Dimensionality of the sample space.
DeviceType	The device type for Kokkos.
Distribution	Type of the distribution to sample from.

Definition at line 26 of file InverseTransformSampling.h.

Member Typedef Documentation

◆ size_type

template<typename T, unsigned Dim, class DeviceType, class Distribution>

using ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::size_type = ippl::detail::size_type

Definition at line 29 of file InverseTransformSampling.h.

◆ view_type

template<typename T, unsigned Dim, class DeviceType, class Distribution>

using ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::view_type = typename ippl::detail::ViewType<Vector<T, Dim>, 1>::view_type

Definition at line 28 of file InverseTransformSampling.h.

Constructor & Destructor Documentation

◆ InverseTransformSampling() [1/3]

template<typename T, unsigned Dim, class DeviceType, class Distribution>

template<class RegionLayout>

ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::InverseTransformSampling	(	Distribution &	dist_r,
		Vector< T, Dim > &	rmax_r,
		Vector< T, Dim > &	rmin_r,
		RegionLayout &	rlayout_r,
		size_type &	ntotal_r )

inline

Constructor for InverseTransformSampling class with domain decomposition.

Parameters

dist_r	The distribution to sample from.
rmax_r	Maximum global range.
rmin_r	Minimum global range.
rlayout_r	The region layout.
ntotal_r	Total number of samples to generate.

Definition at line 45 of file InverseTransformSampling.h.

References ippl::Comm, Dim, dist_m, ntotal_m, and updateBounds().

Here is the call graph for this function:

◆ InverseTransformSampling() [2/3]

template<typename T, unsigned Dim, class DeviceType, class Distribution>

ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::InverseTransformSampling	(	Distribution &	dist_r,
		Vector< T, Dim > &	rmax_r,
		Vector< T, Dim > &	rmin_r,
		Vector< T, Dim > &	locrmax_r,
		Vector< T, Dim > &	locrmin_r,
		size_type &	ntotal_r )

inline

Constructor for InverseTransformSampling class without applying domain decomposition..

Parameters

dist_r	The distribution to sample from.
rmax_r	Maximum global range.
rmin_r	Minimum global range.
locrmax_r	Maximum local (per rank) range.
locrmin_r	Minimum local (per rank) range.
ntotal_	Total number of samples to generate.

Definition at line 74 of file InverseTransformSampling.h.

References dist_m, ntotal_m, and updateBounds().

Here is the call graph for this function:

◆ InverseTransformSampling() [3/3]

template<typename T, unsigned Dim, class DeviceType, class Distribution>

ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::InverseTransformSampling	(	Distribution &	dist_r,
		Vector< T, Dim > &	rmax_r,
		Vector< T, Dim > &	rmin_r,
		size_type &	ntotal_r )

inline

Constructor for InverseTransformSampling class. In this method, we do not consider any domain decomposition.

Parameters

dist_r	The distribution to sample from.
rmax_r	Maximum global range for sampling.
rmin_r	Minimum global range for sampling.
ntotal_r	Total number of samples to generate.

Definition at line 91 of file InverseTransformSampling.h.

References dist_m, nlocal_m, ntotal_m, and updateBounds().

Here is the call graph for this function:

◆ ~InverseTransformSampling()

template<typename T, unsigned Dim, class DeviceType, class Distribution>

ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::~InverseTransformSampling ( )

inline

Deconstructor for InverseTransformSampling class.

Definition at line 166 of file InverseTransformSampling.h.

Member Function Documentation

◆ generate() [1/2]

template<typename T, unsigned Dim, class DeviceType, class Distribution>

void ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::generate	(	view_type	view,
		Kokkos::Random_XorShift64_Pool<>	rand_pool64 )

inline

Generate random samples using inverse transform sampling.

Parameters

view	The view to fill with random samples.
rand_pool64	The random number generator pool.

Definition at line 245 of file InverseTransformSampling.h.

References Dim, dist_m, nlocal_m, umax_m, and umin_m.

◆ generate() [2/2]

template<typename T, unsigned Dim, class DeviceType, class Distribution>

void ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::generate	(	view_type	view,
		size_type	startIndex,
		size_type	endIndex,
		Kokkos::Random_XorShift64_Pool<>	rand_pool64 )

inline

Generate random samples using inverse transform sampling for a specific range of particles.

Parameters

view	The view to fill with random samples.
startIndex	The starting index of view.
endIndex	The ending index of view.
rand_pool64	The random number generator pool.

Definition at line 267 of file InverseTransformSampling.h.

References Dim, dist_m, umax_m, and umin_m.

◆ getLocalSamplesNum()

template<typename T, unsigned Dim, class DeviceType, class Distribution>

KOKKOS_INLINE_FUNCTION size_type ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::getLocalSamplesNum ( ) const

inline

Get the local number of samples.

Returns: The local number of samples.

Definition at line 230 of file InverseTransformSampling.h.

References nlocal_m.

◆ setLocalSamplesNum()

template<typename T, unsigned Dim, class DeviceType, class Distribution>

KOKKOS_INLINE_FUNCTION void ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::setLocalSamplesNum ( size_type nlocal )

inline

Set the local number of particles.

Parameters

nlocal The new number of local particles.

Definition at line 237 of file InverseTransformSampling.h.

References nlocal_m.

◆ updateBounds() [1/2]

template<typename T, unsigned Dim, class DeviceType, class Distribution>

void ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::updateBounds	(	Vector< T, Dim > &	rmax,
		Vector< T, Dim > &	rmin )

inline

Updates the sampling bounds using the CDF without any domain decomposition.

This method allows the user to update the minimum and maximum bounds for the inverse transform sampling method. It recalculates the cumulative distribution function (CDF) values for the new bounds and updates the internal variables to reflect these changes.

Parameters

new_rmax	The new maximum range for sampling. This vector defines the upper bounds for each dimension.
new_rmin	The new minimum range for sampling. This vector defines the lower bounds for each dimension.

Definition at line 155 of file InverseTransformSampling.h.

References Dim, dist_m, umax_m, and umin_m.

◆ updateBounds() [2/2]

template<typename T, unsigned Dim, class DeviceType, class Distribution>

void ippl::random::InverseTransformSampling< T, Dim, DeviceType, Distribution >::updateBounds	(	Vector< T, Dim > &	rmax,
		Vector< T, Dim > &	rmin,
		Vector< T, Dim > &	locrmax,
		Vector< T, Dim > &	locrmin )

inline

Updates the sampling bounds and reinitializes internal variables.

This method allows the user to update the minimum and maximum bounds for the sampling process It recalculates the cumulative distribution function (CDF) values for the new bounds and updates the internal variables to reflect these changes.

Parameters

rmax	The new maximum range for sampling. This vector defines the upper bounds for each dimension.
rmin	The new minimum range for sampling. This vector defines the lower bounds for each dimension.
locrmax	The new local maximum range for sampling. This vector defines the upper bounds for each dimension for a given rank.
locrmin	The new minimum range for sampling. This vector defines the lower bounds for each dimension for a given rank.

Definition at line 117 of file InverseTransformSampling.h.

References ippl::Vector< T, Dim >::begin(), ippl::Comm, Dim, dist_m, ippl::Vector< T, Dim >::end(), nlocal_m, ntotal_m, umax_m, and umin_m.

Referenced by InverseTransformSampling(), InverseTransformSampling(), and InverseTransformSampling().

Here is the call graph for this function:

Here is the caller graph for this function: