#include <fftwaff.h>

Public Types
typedef double	Tsample

typedef std::complex< Tsample >	Tcoeff

typedef aff::Series< Tsample >	Tseries

typedef aff::Series< Tcoeff >	Tspectrum

Public Member Functions
	DRFFTWAFF (const unsigned int &n, const bool &deletearrays=false)

	DRFFTWAFF (const bool &deletearrays=false)

	~DRFFTWAFF ()
	delete plan. More...

Tspectrum	operator() (const Tseries::Tcoc &s, const bool &debug=false) const
	Transform time series to Fourier coefficients. More...

Tseries	operator() (const Tspectrum::Tcoc &s, const bool &debug=false) const
	Transform Fourier coefficients to time series. More...

Tspectrum	operator() (const Tseries::Tcoc &s, const Tsample &dt, const bool &debug=false) const
	Transform time series to Fourier coefficients and scale. More...

Tseries	operator() (const Tspectrum::Tcoc &s, const Tsample &dt, const bool &debug=false) const
	Transform Fourier coefficients to time series and scale. More...

Tsample	scale_series (const Tsample &dt) const
	Return appropriate scaling factor for sampling interval dt. More...

Tsample	scale_spectrum (const Tsample &dt) const
	Return appropriate scaling factor for sampling interval dt. More...

unsigned int	size () const

void	size (const unsigned int &s) const

Static Public Member Functions
static unsigned int	spectrumsize (const unsigned int &n)
	return number of coefficients for given number of samples More...

static unsigned int	seriessize (const unsigned int &n)
	return number of samples for given number of coefficients More...

Private Member Functions
void	create_plan_forward () const
	create plan. More...

void	create_plan_backward () const
	create plan. More...

void	delete_plans () const
	delete plans. More...

void	create_arrays () const
	create plans. More...

void	delete_arrays () const
	delete arrays. More...

unsigned int	ssize () const

void	set_size (const unsigned int &n) const
	prepare FFT settings for size n. More...

Private Attributes
unsigned int	Msize

fftw_plan	Mplan_forward

fftw_plan	Mplan_backward

double *	Mseriesarray

fftw_complex *	Mspectrumarray

Tspectrum	Mspectrum

Tseries	Mseries

bool	Mdeletearrays

Detailed Description

A rigid class to do simple transforms using libdrfftw.a.

uses real double arrays

How to use this class:

You may create one instance of this class and use it to transform several signals in both directions in turn. The class itself takes care of the transform size and creates a new plan if necessary. FFTs are invoked by the bracket operators. You should use the class object like a function. The scaling operators (taking the sampling interval as one of their arguments) return a series or spectrum scaled appropriately to match the values of samples from the corresponding Fourier integral transform (usual convention with $dt$ and $df$ integrals - not $d \omega$ ).

Note: The appropriate number of samples for the time series obtained from a given set of Fourier coefficients is non-unique, if only the Fourier coefficients for positive frequencies are given, as is the case here. A time series with samples and a time series with samples both will results in a set of Fourier coefficients. The method to determine the required number of time samples from the imaginary part of the Nyquist Fourier coefficients only works if the Nyquist coefficient (at least real part) is finite - which is not the case for most of our signals. This way we will obtain Fourier coefficients from time series samples. When reconstructing the time series, we will obtain samples with a sampling interval now being of the original interval. For this reason it is appropriate to set the size of the expected time series explicitely either by using the constructor DRFFTWAFF::DRFFTWAFF(const unsigned int& n) or by calling DRFFTWAFF::size(const unsigned int& s) prior to DRFFTWAFF::operator()(const Tspectrum::Tcoc& s).

See also: Scaling of Fourier transform in FFTW3

Definition at line 129 of file fftwaff.h.

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

Public Types

Public Member Functions

Static Public Member Functions

Private Member Functions

Private Attributes

Detailed Description