Wrapper for a subset of FFTW: FFT's for real functions in one dimension. More...

#include <::c_api>

Classes
class	allocator
	STL allocator replacement using FFTW's memory allocation functions. More...

struct	c_api
	Specialization of c_api for `double` precision. More...

Public Types
typedef T	real
	Real number.

typedef std::complex< T >	complex
	Complex number.

typedef std::vector< real, allocator< real > >	real_vector
	Vector of real numbers using FFTW allocation.

typedef std::vector< complex, allocator< complex > >	complex_vector
	Vector of complex numbers using FFTW allocation.

Static Public Member Functions
static complex_vector &	fft (const real_vector &v, complex_vector &result)
	Computes the real-to-complex Fourier transform of `v` into `result`. More...

static complex_vector::size_type	fft_size (const real_vector &v)
	Returns the output size (exact) for real-to-complex Fourier transform of `v`. More...

static complex_vector	fft (const real_vector &v)
	Computes the real-to-complex Fourier transform of `v`. More...

static real_vector &	ifft (const complex_vector &v, real_vector &result, bool normalize=true)
	Computes the complex-to-real Fourier transform of `v` into `result`. More...

static real_vector::size_type	real_ifft_size (const complex_vector &v)
	Returns the output size (guessed) for complex-to-real Fourier transform of `v`. More...

static real_vector	real_ifft (const complex_vector &v, bool normalize=true)
	Computes the complex-to-real Fourier transform of `v`. More...

Detailed Description

template<typename T>
struct fftw< T >

Wrapper for a subset of FFTW: FFT's for real functions in one dimension.

Specialization of for float precision.

Member Function Documentation

template<typename T>

static complex_vector& fftw< T >::fft	(	const real_vector &	v,
		complex_vector &	result
	)

inlinestatic

Computes the real-to-complex Fourier transform of v into result.

The size of the transform is that of the real component. The expected size of result is v.size() / 2 + 1. Because result is Hermitian, only half of the elements (plus one) are stored.

References fftw< T >::fft_size().

Referenced by fftw< T >::fft().

template<typename T>

static complex_vector fftw< T >::fft ( const real_vector & v )

inlinestatic

Computes the real-to-complex Fourier transform of v.

See also: transform(const real_vector&, complex_vector&)

References fftw< T >::fft(), and fftw< T >::fft_size().

template<typename T>

static complex_vector::size_type fftw< T >::fft_size ( const real_vector & v )

inlinestatic

Returns the output size (exact) for real-to-complex Fourier transform of v.

See also: transform(const real_vector&, complex_vector&)

Referenced by fftw< T >::fft(), and fftw< T >::ifft().

template<typename T>

static real_vector& fftw< T >::ifft	(	const complex_vector &	v,
		real_vector &	result,
		bool	normalize = `true`
	)

inlinestatic

Computes the complex-to-real Fourier transform of v into result.

The size of the transform is that of the real component. The expected size of v is is result.size() / 2 + 1. Because v is Hermitian, only half of the elements (plus one) are considered.

Warning: If normalize is false, the result vector is not normalized; its components must be divided by result.size() for proper normalization.

References fftw< T >::fft_size().

Referenced by fftw< T >::real_ifft(), and LatBuilder::Kernel::RAlpha::valuesVector().

template<typename T>