Content Hotkeys
Nonequispaced fast Fourier transform

Special contributions

Python wrapper for the NFFT library

See the pynfft by G. Vaillant.

For Matlab and Octave

For Linux and macOS, please download nfft-3.4.*.tar.gz and compile with mex interface support (see ./configure --help).
For Windows, ready-to-use mex interfaces can be found on the download page.

For Mathematica

A Mathematica implementation for performing the NFFT as well as a Mathematica notebook demonstrating the use of the package written by Sseziwa Mukasa (mukasa(AT) can be downloaded.

For Julia

A Julia implementation for performing the NFFT written by Tobias Knopp see.

For Windows & Visual C

The NFFT2.0 is also available for Windows as Visual C project by Eskofier Bjoern (bjoern.eskofier(AT)
Using the NFFT3.0 for Windows see the FAQ.


See the CUNFFT - Nonequispaced FFT and it's inversion in CUDA on the homepage of Stefan Kunis.


Methods, similar to the NFFT, are known as

  • unequally-spaced fast Fourier transform (USFFT by G. Beylkin).
  • generalized fast Fourier transform (GFFT by A. Dutt and V. Rokhlin),
  • nonuniform fast Fourier transform (NUFFT by J.A. Fessler and B.P. Sutton),
  • nonequispaced fast Fourier transform NFFT by K. Fourmount),
  • irregularly spaced data analysis (ISDA by J. Pelt)