In this repository we include the data and the code for the generation of labelled temporal networks with the LETN method, presented in this pre-print.
Our method takes as input a temporal network, and looks at local patterns, grouping patterns of nodes with the same attribute, thus resulting in a generation that takes into account communities when the attribute is the community the nodes belong to.
This is particularly relevant when producing surrogate networks to study face-to-face interactions. For this reason we tested our method with the SocioPatterns datasets.
In this repo we include the notebook Usage.ipynb, in which we show how to use our module. You can plug-in your dataset by simply changing the file names. In our files we include the edges in the following format: timestamp node1 node2. The files with the metadata (if available) have the following format: node metadata.
If you find the paper and this code useful, please cite us (journal publication under review) as:
@article{girardini2025community,
title={Community Aware Temporal Network Generation},
author={Girardini, Nicol{\`o} Alessandro and Longa, Antonio and Trebucchi, Gaia and Cencetti, Giulia and Passerini, Andrea and Lepri, Bruno},
journal={arXiv preprint arXiv:2501.07327},
year={2025},
doi={https://doi.org/10.48550/arXiv.2501.07327}
}