martingerlach/testing-statistical-laws-in-complex-systems

testing-statistical-laws-in-complex-systems

This repository contains the data and the code for the paper.

Martin Gerlach, Eduardo G. Altmann, Testing Statistical Laws in Complex Systems, Physical Review Letters, to appear (2019).

Setup:

• get the repository:

git clone https://github.com/martingerlach/testing-statistical-laws-in-complex-systems

• install the following python packages (e.g. via pip):

numpy, scipy, pandas, matplotlib, mpmath, statsmodels

Navigation

• The folder code/ contains notebooks with the analysis of each Dataset (.ipynb or .html)
  • Earthquake data: Analysis_real-data_earthquakes.ipynb
  • Text Interevent data: Analysis_real-data_texts-interevent-times.ipynb
  • Text Frequency-rank data: Analysis_real-data_texts-rank-frequency.ipynb
  • Network data: Analysis_real-data_networks-degree.ipynb
  • Synthetic data: Analysis_synthetic-data.ipynb
• The code for all functions can be found in src/

Datasets

We analyze several datasets which we included in the repo. They are contained in the folder data/

• Earthquakes: from the Southern California Earthquake Data Center
• Books from Project Gutenberg: from the Standardized Project Gutenberg Corpus
• Networks: KONECT Project: Internet Topology in the folder data/networks/. The timeseries data for the sequence of degrees is in the folder data/networks/sampling and was obtained using the code in src/sampling-nets.py
• Synthetic data: The timeseries can be generated using the code in src/modules_mcmc_zipf.py. There is an example of a correlated timeseries in data/synthetic/ts_synthetic_Ntypes1000_Ntokens100000_alpha1.5_mu0.01_k5