By Cleve Moler, MathWorks
The fft function in MATLAB® 6 (released in 2000) and subsequent versions uses fast algorithms for any length, even a prime. Earlier versions of MATLAB used fast algorithms only when the length was a product of small primes.
In this article, originally published in 1992, Cleve Moler uses plots to explain the behavior of the finite Fourier transform and the fast finite Fourier transform algorithms implemented in MATLAB at that time.
Published 1992 - 92034v00