This repository contains the codes to generate different types of constrained (and other) power-law surrogates. It allows you to
- generate constrained surrogates based on a time series
- perform simple hypothesis tests with constrained surrogates
- reproduce the results of the manuscript Nonparametric power-law surrogates, by Jack Murdoch Moore, Gang Yan, and Eduardo G. Altmann
A tutorial to generate surrogates based on a new or existing time series is given in the Jupyter notebook 'tutorial.ipynb' in the current folder.
In order to reproduce the results of the manuscript, you should run the notebook 'generate-results.ipynb' with the parameters of the manuscript (to generate the results) and the notebook 'make-figures.ipynb' (to generate the figures). Both 'generate-results.ipynb' and 'make-figures.ipynb' are in the folder reproduce-paper.
- src: contains source code (i.e., the module 'constrained_power_law_surrogates.py')
- time-series: contains the data used in this repository
- reproduce-paper: code, output data and figures that reproduce the results of the manuscript
- 'requirements.txt': python packages required in the repository.
- 'tutorial.ipynb': A tutorial to generate surrogates based on a new or existing time series.
- "Nonparametric Power-Law Surrogates", Jack Murdoch Moore, Gang Yan, and Eduardo G. Altmann, Phys. Rev. X 12, 021056 (2022)