A wrapper for FFTW plans for fast Fourier transforms. More...

#include <FourierTransform.h>

Inheritance diagram for ndarray::FourierTransform< T, N >:

Public Types
typedef boost::shared_ptr< FourierTransform >	Ptr

typedef detail::FourierTraits< T >::ElementX	ElementX
	Real-space array data type;.

typedef detail::FourierTraits< T >::ElementK	ElementK
	Fourier-space array data type;.

typedef Vector< Size, N >	Index
	Shape type for arrays.

typedef Array< ElementX, N, N >	ArrayX
	Real-space array type.

typedef Array< ElementK, N, N >	ArrayK
	Fourier-space array type.

typedef Vector< Size, N+1 >	MultiplexIndex
	Shape type for multiplexed arrays.

typedef Array< ElementX, N+1, N+1 >	MultiplexArrayX
	Real-space multiplexed array type.

typedef Array< ElementK, N+1, N+1 >	MultiplexArrayK
	Fourier-space multiplexed array type.

Public Member Functions
void	execute ()
	Execute the FFTW plan.

Static Public Member Functions
static Ptr	planForward (Index const &shape, ArrayX &x, ArrayK &k)
	Create a plan for forward-transforming a single N-dimensional array.

static Ptr	planInverse (Index const &shape, ArrayK &k, ArrayX &x)
	Create a plan for inverse-transforming a single N-dimensional array.

static Ptr	planMultiplexForward (MultiplexIndex const &shape, MultiplexArrayX &x, MultiplexArrayK &k)
	Create a plan for forward-transforming a sequence of nested N-dimensional arrays.

static Ptr	planMultiplexInverse (MultiplexIndex const &shape, MultiplexArrayK &k, MultiplexArrayX &x)
	Create a plan for inverse-transforming a sequence of nested N-dimensional arrays.

template<int M>
static Array< ElementX, M, M >	initializeX (Vector< Size, M > const &shape)
	Create a new real-space array with the given real-space shape.

template<int M>
static Array< ElementK, M, M >	initializeK (Vector< Size, M > const &shape)
	Create a new Fourier-space array with the given real-space shape.

template<int M>
static void	initialize (Vector< Size, M > const &shape, Array< ElementX, M, M > &x, Array< ElementK, M, M > &k)
	Initialize, as necessary, a pair of arrays with the given real-space shape.

Detailed Description

template<typename T, int N>
class ndarray::FourierTransform< T, N >

A wrapper for FFTW plans for fast Fourier transforms.

An instance of FourierTransform holds an FFTW "plan", providing repeated forward or inverse FFTs of predetermined arrays.

Multiplex plans can also be generated to perform an N-dimensional FFT on the nested arrays of an (N+1)-dimensional array.

Static member functions of FourierTransform are used to create instances, and optionally initialize the involved arrays.

Member Function Documentation

◆ initialize()

template<typename T , int N>

template<int M>

static void ndarray::FourierTransform< T, N >::initialize	(	Vector< Size, M > const &	shape,
		Array< ElementX, M, M > &	x,
		Array< ElementK, M, M > &	k
	)

static

Initialize, as necessary, a pair of arrays with the given real-space shape.

If either array is not empty, it must be consistent with the given shape.

◆ planForward()

template<typename T , int N>

static Ptr ndarray::FourierTransform< T, N >::planForward	(	Index const &	shape,
		ArrayX &	x,
		ArrayK &	k
	)

static

Create a plan for forward-transforming a single N-dimensional array.

Arrays will be initialized with new memory if empty. If they are not empty, existing data may be overwritten when the plan is created.

Parameters

shape	Shape of the real-space array.
x	Input real-space array.
k	Output Fourier-space array.

◆ planInverse()

template<typename T , int N>

static Ptr ndarray::FourierTransform< T, N >::planInverse	(	Index const &	shape,
		ArrayK &	k,
		ArrayX &	x
	)

static

Create a plan for inverse-transforming a single N-dimensional array.

Arrays will be initialized with new memory if empty. If they are not empty, existing data may be overwritten when the plan is created.

Parameters

shape	Shape of the real-space array.
k	Input Fourier-space array.
x	Output real-space array.

◆ planMultiplexForward()

template<typename T , int N>

static Ptr ndarray::FourierTransform< T, N >::planMultiplexForward	(	MultiplexIndex const &	shape,
		MultiplexArrayX &	x,
		MultiplexArrayK &	k
	)

static

Create a plan for forward-transforming a sequence of nested N-dimensional arrays.

Arrays will be initialized with new memory if empty. If they are not empty, existing data may be overwritten when the plan is created.

Parameters

shape	Shape of the real-space array. First dimension is multiplexed.
x	Input real-space array.
k	Output Fourier-space array.

◆ planMultiplexInverse()

template<typename T , int N>

static Ptr ndarray::FourierTransform< T, N >::planMultiplexInverse	(	MultiplexIndex const &	shape,
		MultiplexArrayK &	k,
		MultiplexArrayX &	x
	)

static

Create a plan for inverse-transforming a sequence of nested N-dimensional arrays.

Arrays will be initialized with new memory if empty. If they are not empty, existing data may be overwritten when the plan is created.

Parameters

shape	Shape of the real-space array. First dimension is multiplexed.
k	Input Fourier-space array.
x	Output real-space array.

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

/home/runner/work/ndarray/ndarray/include/ndarray/fft/FourierTransform.h

Public Types

Public Member Functions

Static Public Member Functions

Detailed Description

Member Function Documentation

◆ initialize()

◆ planForward()

◆ planInverse()

◆ planMultiplexForward()

◆ planMultiplexInverse()