pulsesuite.core.fftw

Attributes

fft_1d
ifft_1d
fftc_1d
ifftc_1d
fft_2d
ifft_2d
fftc_2d
ifftc_2d
fft_3d
ifft_3d
fftc_3d
ifftc_3d

Functions

nyquist_1D(→ None)
nyquist_2D(→ None)
nyquist_3D(→ None)
fft_1D(→ None)
ifft_1D(→ None)
fftc_1D(→ None)
ifftc_1D(→ None)
fft_2D(→ None)
ifft_2D(→ None)
fftc_2D(→ None)
ifftc_2D(→ None)
fftw_initialize_2D(→ None)	API-compat stub: planning handled by SciPy/pyFFTW internally.
fftw_initialize_3D(→ None)	API-compat stub: planning handled by SciPy/pyFFTW internally.
fft_3D(→ None)
ifft_3D(→ None)
fftc_3D(→ None)
ifftc_3D(→ None)
CreateHT(→ None)	Precompute Hankel transform matrix HT(Nr, Nr) from J0 zeros.
HankelTransform(→ None)	Apply HT @ f in-place (f length must match HT dimension).
Transform(→ None)	If Z has shape (1, Nr, Nt), apply Hankel along r and FFT along t; else FFT3D.
iTransform(→ None)	Inverse of Transform using ifft along t; Hankel is its own inverse for this discretization.

Module Contents

pulsesuite.core.fftw.nyquist_1D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.nyquist_2D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.nyquist_3D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.fft_1D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.ifft_1D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.fftc_1D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.ifftc_1D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.fft_2D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.ifft_2D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.fftc_2D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.ifftc_2D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.fftw_initialize_2D(Z: numpy.typing.NDArray[_dc]) → None

API-compat stub: planning handled by SciPy/pyFFTW internally.

pulsesuite.core.fftw.fftw_initialize_3D(Z: numpy.typing.NDArray[_dc]) → None

API-compat stub: planning handled by SciPy/pyFFTW internally.

pulsesuite.core.fftw.fft_3D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.ifft_3D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.fftc_3D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.ifftc_3D(Z: numpy.typing.NDArray[_dc]) → None

pulsesuite.core.fftw.CreateHT(Nr: int) → None

Precompute Hankel transform matrix HT(Nr, Nr) from J0 zeros.

Fortran reference:

HT(m,n) = (2/a(Nr+1)) * J0( a(m)*a(n) / a(Nr+1) ) / J1(a(n))**2, m,n=1..Nr

where a(:) are the first Nr+1 zeros of J0.

We store a length-(Nr+1) array and a (Nr,Nr) HT in double precision.

pulsesuite.core.fftw.HankelTransform(f: numpy.typing.NDArray[_dc]) → None

Apply HT @ f in-place (f length must match HT dimension).

pulsesuite.core.fftw.Transform(Z: numpy.typing.NDArray[_dc]) → None

If Z has shape (1, Nr, Nt), apply Hankel along r and FFT along t; else FFT3D.

Fortran logic:

if size(Z,1) == 1:

for k: HankelTransform(Z(1,:,k)) for j: FFT(Z(1,j,:))

else:

FFT(Z)

pulsesuite.core.fftw.iTransform(Z: numpy.typing.NDArray[_dc]) → None

Inverse of Transform using ifft along t; Hankel is its own inverse for this discretization.

Fortran logic called HankelTransform both ways; we mirror that.

pulsesuite.core.fftw.fft_1d

pulsesuite.core.fftw.ifft_1d

pulsesuite.core.fftw.fftc_1d

pulsesuite.core.fftw.ifftc_1d

pulsesuite.core.fftw.fft_2d

pulsesuite.core.fftw.ifft_2d

pulsesuite.core.fftw.fftc_2d

pulsesuite.core.fftw.ifftc_2d

pulsesuite.core.fftw.fft_3d

pulsesuite.core.fftw.ifft_3d

pulsesuite.core.fftw.fftc_3d

pulsesuite.core.fftw.ifftc_3d