1.1 - 8.10 - FFT (ML Engine) - Teradata Vantage

Teradata Vantage™ - Machine Learning Engine Analytic Function Reference

Teradata Vantage
Release Number
October 2019
Content Type
Programming Reference
Publication ID
English (United States)

The FFT function uses a Fast Fourier Transform (FFT) algorithm to compute the discrete Fourier Transform (DFT) of each signal in one or more input table columns. A signal can be either real or complex, and can have one, two, or three dimensions. If the signal length is not a power of two, the function either pads or truncates it to the closest power of two.

This is the DFT of a time sequence of length N, 0..N-1:
X(k) = Xk(N - k)

where k ϵ 0..N-1.

  • The FFT of a time sequence of length 1 is the one-element sequence itself.
  • The FFT of a time sequence of length 2 has only real values.
  • The FFT of a time sequence of length 4 or greater has conjugate symmetry.

To recover the original signals, use the IFFT (ML Engine) function.