Scribd
Upload
159 views

Scipy - FFT Cheat Sheet

This document provides a cheat sheet for using the scipy.fft module for Fourier transforms in Python. It lists the main functions for computing the discrete Fourier transform, discrete cosine transform, and discrete sine transform on real and complex 1D, 2D, and nD data. It also describes some additional useful functions for shifting the outputs, getting frequencies, and setting the number of worker threads. In the last section, it briefly explains the differences between the discrete Fourier transform, discrete cosine transform, and discrete sine transform.

Uploaded by

Asif Raza
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
159 views

Scipy - FFT Cheat Sheet

This document provides a cheat sheet for using the scipy.fft module for Fourier transforms in Python. It lists the main functions for computing the discrete Fourier transform, discrete cosine transform, and discrete sine transform on real and complex 1D, 2D, and nD data. It also describes some additional useful functions for shifting the outputs, getting frequencies, and setting the number of worker threads. In the last section, it briefly explains the differences between the discrete Fourier transform, discrete cosine transform, and discrete sine transform.

Uploaded by

Asif Raza
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

scipy.

fft Cheat Sheet


realpython.com

scipy.fft Cheat Sheet


To use the scipy.fft module, you need to remember a lot of math and function naming conventions. This cheat sheet contains some helpful tips in a condensed format, which
should make using the library easier.

New to the scipy.fft module? Check out Fourier Transforms With scipy.fft: Python Signal Processing.

Note: Functions with names beginning with i are the inverse of the functions that share the name without the i.

fft(), fft2(), fftn() dct(), dctn()

Compute the fast Fourier transform on one-, two-, and n-dimensional input. Compute the discrete cosine transform on one- and n-dimensional input. Real-
Work on real or complex input. valued input and output. Often a better choice than fft() or rfft().

rfft(), rfft2(), rfftn() dst(), dstn()

Compute the fast Fourier transform on one-, two-, and n-dimensional real input. Compute the discrete sine transform on one- and n-dimensional input. Real-
Faster than the fft*() functions. valued input and output.

hfft(), hfft2(), hfftn() fftshift()

Compute the fast Fourier transform on one-, two-, and n-dimensional input that By default, the fft and fftfreq functions return all the positive components,
has Hermitian symmetry. Produce real-valued output. followed by all the negative components. This function swaps the two halves so
that the zero-frequency component is in the center: [0, 1, 2, -3, -2, -1] →
[-3, -2, -1, 0, 1, 2, 3].

realpython.com 1
scipy.fft Cheat Sheet

fftfreq(), rfftfreq() set_workers()

Get the frequencies of each bin returned by the FFT functions, such as the X-axis. Context manager to set the number of parallel worker threads used to compute
the FFT.

next_fast_len()
get_workers()
The FFT functions work fastest at certain lengths of input. This function returns
the next largest one for use with zero-padding. Returns the number of workers used in the current context.

Note: I’ve left out backend-switching functions since, at the time of writing, SciPy ships with only one available backend.

DFT vs DCT vs DST

There are three main transforms that scipy.fft deals with: the discrete Fourier transform, the discrete cosine transform, and the discrete sine transform. This section is a quick
reminder of what each transform is.

Discrete Fourier transform (DFT):

• Converts between a signal’s time-domain representation and its frequency spectrum


• Uses both cosine and sine waves to decompose the signal into frequency components
• Handles complex input and output
• Assumes a function repeats to infinity

Discrete cosine transform (DCT):

• Does the same job as the DFT


• Uses only cosines to decompose the function
• Both input and output are real-valued
• Performs faster than the DFT
• Assumes a function has even symmetry (symmetry about the Y-axis), which extends to infinity

Discrete sine transform (DST):

• Does the same job as the DFT and the DST


• Uses sines to decompose the function
• Both input and output are real-valued
• Performs faster than the DFT
• Assumes a function has odd symmetry (symmetry about the origin) that extends to infinity
• May introduce high-frequency components if input data doesn’t have odd symmetry

realpython.com 2

You might also like