Represents a Scaled TruncatedNormal (Gaussian) probability distribution. More...

#include <prng_extension.hpp>

Collaboration diagram for MC::Distributions::ScaledTruncatedNormal< F >:

Public Member Functions
constexpr	ScaledTruncatedNormal (F factor, F m, F s, F l, F u)

template<class DeviceType >
KOKKOS_INLINE_FUNCTION F	draw (Kokkos::Random_XorShift1024< DeviceType > &gen) const

KOKKOS_INLINE_FUNCTION F	mean () const

KOKKOS_INLINE_FUNCTION F	var () const

KOKKOS_INLINE_FUNCTION F	skewness () const

Static Public Member Functions
template<class DeviceType >
static KOKKOS_INLINE_FUNCTION F	draw_from (Kokkos::Random_XorShift1024< DeviceType > &gen, F factor, F mu, F sigma, F lower, F upper)

Public Attributes
F	scale_factor

F	inverse_factor

TruncatedNormal< F >	dist

Detailed Description

template<FloatingPointType F>
struct MC::Distributions::ScaledTruncatedNormal< F >

Represents a Scaled TruncatedNormal (Gaussian) probability distribution.

The normal distribution is parameterized by a mean, a standard deviation a scale factor and lower and upper bound. It supports random sampling, computing statistical properties. To sample from [a, b] where |a - b| << 1, methods perform better if we draw from xf = [a * factor, b * factor] and then scale back using x = xf / factor.

Template Parameters

F	Floating-point type (must satisfy `FloatingPointType`).

Constructor & Destructor Documentation

◆ ScaledTruncatedNormal()

template<FloatingPointType F>

MC::Distributions::ScaledTruncatedNormal< F >::ScaledTruncatedNormal	(	F	factor,
		F	m,
		F	s,
		F	l,
		F	u )

inlineconstexpr

Member Function Documentation

◆ draw()

template<FloatingPointType F>

template<class DeviceType >

KOKKOS_INLINE_FUNCTION F MC::Distributions::ScaledTruncatedNormal< F >::draw ( Kokkos::Random_XorShift1024< DeviceType > & gen ) const

inline

◆ draw_from()

template<FloatingPointType F>

template<class DeviceType >

static KOKKOS_INLINE_FUNCTION F MC::Distributions::ScaledTruncatedNormal< F >::draw_from	(	Kokkos::Random_XorShift1024< DeviceType > &	gen,
		F	factor,
		F	mu,
		F	sigma,
		F	lower,
		F	upper )