alpaka::rand::distribution::NormalReal< T_Result > Struct Template Reference

Used to sample floating-point values from a normal(-/gaussian) distribution. More...

#include <NormalReal.hpp>

Public Types
using	result_type = T_Result

Public Member Functions
constexpr	NormalReal (T_Result mean=T_Result{0}, T_Result stdDev=T_Result{1})
	Constructs normal(-/gaussian) distribution with given parameters.
template<concepts::UniformStdEngine T_Engine>
constexpr result_type	operator() (T_Engine &engine)
	Selects a value from a normal (-/gaussian) distribution for the configured (mean,stdDev) settings.

Private Types
using	T_Impl = internal::BoxMuller<T_Result>

Private Attributes
T_Impl	m_impl
result_type const	m_mean
result_type const	m_stdDev

Detailed Description

template<std::floating_point T_Result>
struct alpaka::rand::distribution::NormalReal< T_Result >

Used to sample floating-point values from a normal(-/gaussian) distribution.

Usage is analogous to std::normal_distribution<T_Result>

See also

https://en.cppreference.com/w/cpp/numeric/random/normal_distribution.html -> Generates N(mean, stddev). Is using the Box-Muller method of generating normally distributed random numbers from a uniform distribution.

Note

: currently supports 32 and 64 bit floating point types.

Definition at line 73 of file NormalReal.hpp.

Member Typedef Documentation

result_type

template<std::floating_point T_Result>

using alpaka::rand::distribution::NormalReal< T_Result >::result_type = T_Result

Definition at line 75 of file NormalReal.hpp.

T_Impl

template<std::floating_point T_Result>

using alpaka::rand::distribution::NormalReal< T_Result >::T_Impl = internal::BoxMuller<T_Result>

private

Definition at line 112 of file NormalReal.hpp.

Constructor & Destructor Documentation

NormalReal()

template<std::floating_point T_Result>

alpaka::rand::distribution::NormalReal< T_Result >::NormalReal ( T_Result mean = T_Result{0}, T_Result stdDev = T_Result{1} )

inlineexplicitconstexpr

Constructs normal(-/gaussian) distribution with given parameters.

Parameters

mean	Mean of the target normal distribution.
stdDev	Standard deviation of the target normal distribution.

The default is N(0,1). The distribution is sampled using the Box-Muller method. This implementation keeps an internal state, therefore each thread/worker must use its own instance to avoid data races when the same object is accessed concurrently by multiple workers.

Usage is otherwise analogous to std::normal_distribution<T_Result>

See also

https://en.cppreference.com/w/cpp/numeric/random/normal_distribution.html

Definition at line 92 of file NormalReal.hpp.

Member Function Documentation

operator()()

template<std::floating_point T_Result>

template<concepts::UniformStdEngine T_Engine>

result_type alpaka::rand::distribution::NormalReal< T_Result >::operator() ( T_Engine & engine )

inlineconstexpr

Selects a value from a normal (-/gaussian) distribution for the configured (mean,stdDev) settings.

Parameters

engine

a random engine conforming to the UniformRandomEngine concept (currently accepts stdlib uniform engines and alpaka engines included in the alpaka::rand::engine namespace)

Returns

a floating-point value sampled from the configured distribution.

Definition at line 105 of file NormalReal.hpp.

Member Data Documentation

m_impl

template<std::floating_point T_Result>

T_Impl alpaka::rand::distribution::NormalReal< T_Result >::m_impl

private

Definition at line 114 of file NormalReal.hpp.

m_mean

template<std::floating_point T_Result>

result_type const alpaka::rand::distribution::NormalReal< T_Result >::m_mean

private

Definition at line 115 of file NormalReal.hpp.

m_stdDev

template<std::floating_point T_Result>

result_type const alpaka::rand::distribution::NormalReal< T_Result >::m_stdDev

private

Definition at line 116 of file NormalReal.hpp.

---

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

• /home/docs/checkouts/readthedocs.org/user_builds/alpaka3/checkouts/latest/include/alpaka/rand/distribution/NormalReal.hpp