Eventgraphs

A Python library for analysing sequences of event-based data and temporal networks.

Features

1. Build EventGraphs (static representation of a temporal network) using arbitrary joining rules.
2. Calculate inter-event time distributions (including motif-dependent inter-event times).
3. Calculate temporal motif distributions.
4. Network decomposition into temporal components.
5. Calculate event centralities.
6. Dimension-reduction and component clustering.
7. Plot network components and EventGraphs.
8. IO functionality (saving as JSON).

Installation

For the latest version, installation from Github is recommended. S The PyPI package is updated periodically.

Install from Github (latest version, recommended)

pip install git+https://github.com/empiricalstateofmind/eventgraphs

Install from PyPI

pip install eventgraphs

Getting Started

The best place to get started using EventGraphs is with the tutorial here.

References

Event Graphs: Advances and Applications of Second-Order Time-Unfolded Temporal Network Models. Andrew Mellor (2018) [ArXiv]

Analysing Collective Behaviour in Temporal Networks Using Event Graphs and Temporal Motifs. Andrew Mellor (2018) [ArXiv]

The Temporal Event Graph. Andrew Mellor (2017) [Journal of Complex Networks] [ArXiv]

Please consider citing these papers if you use this code for further research.