1.1 - 8.10 - Fast Fourier Transform Functions (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)

Fast Fourier Transform (FFT), developed by Cooley and Tukey in 1965, is an algorithm that computes the discrete Fourier Transform (DFT) of a signal. FFT significantly reduces the complexity of the Fourier Transform algorithm by exploiting the symmetry and periodicity of a Fourier Transform and using a divide-and-conquer strategy.

The divide-and conquer-strategy that ML Engine FFT function uses is Radix-2, Radix-4, or Radix-8, for a signal whose length is a power of 2, 4, or 8, respectively.

Function Description
FFT (ML Engine) Uses FFT algorithm to compute DFT of each signal in one or more input table columns.
IFFT (ML Engine) Uses inverse Fast Fourier Transform (IFFT) algorithm (also called a Fourier synthesis algorithm) to reverse Fast Fourier Transform performed by FFT function; that is, the IFFT function takes a frequency domain representation and combines the contributions of all the different frequencies to recover the original signal.