mgcpp::cufft Namespace Reference

Enumerations
enum	fft_direction { fft_direction::forward = CUFFT_FORWARD, fft_direction::inverse = CUFFT_INVERSE }

Functions
outcome::result< void >	rfft (size_t n, float const *x, cuComplex *result)
outcome::result< void >	rfft (size_t n, double const *x, cuDoubleComplex *result)
outcome::result< void >	irfft (size_t n, cuComplex const *x, float *result)
outcome::result< void >	irfft (size_t n, cuDoubleComplex const *x, double *result)
outcome::result< void >	cfft (size_t n, cuComplex const *x, cuComplex *result, fft_direction direction)
outcome::result< void >	cfft (size_t n, cuDoubleComplex const *x, cuDoubleComplex *result, fft_direction direction)
outcome::result< void >	rfft2 (size_t n, size_t m, float const *x, cuComplex *result)
outcome::result< void >	rfft2 (size_t n, size_t m, double const *x, cuDoubleComplex *result)
outcome::result< void >	irfft2 (size_t n, size_t m, cuComplex const *x, float *result)
outcome::result< void >	irfft2 (size_t n, size_t m, cuDoubleComplex const *x, double *result)
outcome::result< void >	cfft2 (size_t n, size_t m, cuComplex const *x, cuComplex *result, fft_direction direction)
outcome::result< void >	cfft2 (size_t n, size_t m, cuDoubleComplex const *x, cuDoubleComplex *result, fft_direction direction)

Enumeration Type Documentation

fft_direction

enum mgcpp::cufft::fft_direction

strong

Enumerator
forward	Forward FFT (time domain -> frequency domain representation)
inverse	Inverse FFT (frequency domain -> time domain representation)

Function Documentation

cfft() [1/2]

outcome::result<void> mgcpp::cufft::cfft	(	size_t	n,
		cuComplex const *	x,
		cuComplex *	result,
		fft_direction	direction
	)

Performs single-precision complex-to-complex FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fft size
x	input array of n complex values
result	the fft result, which is an array of n complex values.
direction	cufft::fft_direction::forward for forward FFT, or cufft::fft_direction::inverse to perform an inverse (unnormalized) FFT.

cfft() [2/2]

outcome::result<void> mgcpp::cufft::cfft	(	size_t	n,
		cuDoubleComplex const *	x,
		cuDoubleComplex *	result,
		fft_direction	direction
	)

Performs double-precision complex-to-complex FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fft size
x	input array of n complex values
result	the fft result, which is an array of n complex values.
direction	cufft::fft_direction::forward for forward FFT, or cufft::fft_direction::inverse to perform an inverse (unnormalized) FFT.

cfft2() [1/2]

outcome::result<void> mgcpp::cufft::cfft2	(	size_t	n,
		size_t	m,
		cuComplex const *	x,
		cuComplex *	result,
		fft_direction	direction
	)

Performs two-dimensional single-precision complex-to-complex FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fastest-changing dimension
m	slowest-changing dimension
x	input array of n x m interleaved complex values
result	the fft result, which is an array of n x m real numbers.
direction	cufft::fft_direction::forward for forward FFT, or cufft::fft_direction::inverse to perform an inverse (unnormalized) FFT.

cfft2() [2/2]

outcome::result<void> mgcpp::cufft::cfft2	(	size_t	n,
		size_t	m,
		cuDoubleComplex const *	x,
		cuDoubleComplex *	result,
		fft_direction	direction
	)

Performs two-dimensional double-precision complex-to-complex FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fastest-changing dimension
m	slowest-changing dimension
x	input array of n x m interleaved complex values
result	the fft result, which is an array of n x m real numbers.
direction	cufft::fft_direction::forward for forward FFT, or cufft::fft_direction::inverse to perform an inverse (unnormalized) FFT.

irfft() [1/2]

outcome::result<void> mgcpp::cufft::irfft	(	size_t	n,
		cuComplex const *	x,
		float *	result
	)

Performs single-precision complex-to-real inverse unnormalized FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fft size
x	input array of floor(n/2)+1 interleaved complex values
result	the fft result, which is an array of n real numbers.

irfft() [2/2]

outcome::result<void> mgcpp::cufft::irfft	(	size_t	n,
		cuDoubleComplex const *	x,
		double *	result
	)

Performs double-precision complex-to-real inverse unnormalized FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fft size
x	input array of floor(n/2)+1 interleaved complex values
result	the fft result, which is an array of n real numbers.

irfft2() [1/2]

outcome::result<void> mgcpp::cufft::irfft2	(	size_t	n,
		size_t	m,
		cuComplex const *	x,
		float *	result
	)

Performs two-dimensional single-precision complex-to-real inverse unnormalized FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fastest-changing dimension
m	slowest-changing dimension
x	input array of (floor(n/2)+1) x m interleaved complex values
result	the fft result, which is an array of n x m real numbers.

irfft2() [2/2]

outcome::result<void> mgcpp::cufft::irfft2	(	size_t	n,
		size_t	m,
		cuDoubleComplex const *	x,
		double *	result
	)

Performs two-dimensional double-precision complex-to-real inverse unnormalized FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fastest-changing dimension
m	slowest-changing dimension
x	input array of (floor(n/2)+1) x m interleaved complex values
result	the fft result, which is an array of n x m real numbers.

rfft() [1/2]

outcome::result<void> mgcpp::cufft::rfft	(	size_t	n,
		float const *	x,
		cuComplex *	result
	)

Performs single-precision real-to-complex forward FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fft size
x	input array of n real values
result	the fft result, which is an array of floor(n/2)+1 complex numbers.

rfft() [2/2]

outcome::result<void> mgcpp::cufft::rfft	(	size_t	n,
		double const *	x,
		cuDoubleComplex *	result
	)

Performs double-precision real-to-complex forward FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fft size
x	input array of n real values
result	the fft result, which is an array of floor(n/2)+1 complex numbers.

rfft2() [1/2]

outcome::result<void> mgcpp::cufft::rfft2	(	size_t	n,
		size_t	m,
		float const *	x,
		cuComplex *	result
	)

Performs two-dimensional single-precision real-to-complex forward FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fastest-changing dimension
m	slowest-changing dimension
x	input array of (n x m) real values
result	the fft result, which is an array of floor(n/2)+1 complex numbers.

rfft2() [2/2]

outcome::result<void> mgcpp::cufft::rfft2	(	size_t	n,
		size_t	m,
		double const *	x,
		cuDoubleComplex *	result
	)

Performs two-dimensional double-precision real-to-complex forward FFT. Effectively calls the corresponding CuFFT function.

Parameters

n	fastest-changing dimension
m	slowest-changing dimension
x	input array of (n x m) real values
result	the fft result, which is an array of (floor(n/2)+1) x m complex numbers.