Iterated Amplitude Adjusted Fourier Transform time series surrogates

This repo implements that iterative amplitude adjusted Fourier transform method to generate time series surrogates (i.e randomised copies of a given time series) which preserve the power spectrum (and consequently the autocorrelation) of the original series. For a detailed discussion on the method, we refer to the paper by Venema, Ament & Simmer (2006) below.

Installation & Usage

• Installation: The algorithm is implemented as a Python module with the idea that you can simply download iaaft.py in your project directory and import is a Python module with import iaaft.

• Prequisite Python packages

  • numpy
  • tqdm (progress bar display)

• Usage: Call iaaft.surrogates() with appropriate arguments.

• Example: The script example.py contains a simple example with an autoregressive time series of order 1 is given as in put and a total of 1000 IAAFT surrogates are generated (results shown in figure above). To run this script, it takes around 35 secs on a Intel® Core™ i9-9880H CPU @ 2.30GHz.

References

Venema, V., Ament, F. & Simmer, C. A stochastic iterative amplitude adjusted Fourier Transform algorithm with improved accuracy (2006), Nonlin. Proc. Geophys. 13, pp. 321--328
https://doi.org/10.5194/npg-13-321-2006

TODO

• First working implementation
• Optimize code for faster performance
• Documentation
• Example
• Improve README (installation, license, usage, etc.)

License

• Copyright © Bedartha Goswami.

Issues?

If you find any issues simply open a bug report, or send an email to [email protected]