Higher-order Analytics with pathpy
This tutorial gives an in-depth introduction to the python Open Source data analytics package pathpy
.
pathpy
provides a new approach to analyse and visualise time-series data on complex networks. Examples include time-stamped social networks, temporal proximitydata, traces on flow processes in networks, passenger itineraries in transportation networks, user clickstreams in the Web, citation networks, or biological pathway data. Building on the higher- and multi-order statistical modelling framework introduced in [1] and [2], pathpy
offers machine learning techniques to select optimal, higher-order network models for your data. It then uses these models to detect, model, and visualise higher-order dependencies and patterns discarded by state-of-the-art data analytics methods that focus on dyadic links.
The following video gives a high-level explanation of the science behind pathpy
:
Details of the scientific background can be found in the following published works:
- I Scholtes: When is a network a network? Multi-Order Graphical Model Selection in Pathways and Temporal Networks, In KDD’17 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, Nova Scotia, Canada, August 13-17, 2017
- I Scholtes, N Wider, A Garas: Higher-Order Aggregate Networks in the Analysis of Temporal Networks: Path structures and centralities, The European Physical Journal B, 89:61, March 2016
- I Scholtes, N Wider, R Pfitzner, A Garas, CJ Tessone, F Schweitzer: Causality-driven slow-down and speed-up of diffusion in non-Markovian temporal networks, Nature Communications, 5, September 2014
- R Pfitzner, I Scholtes, A Garas, CJ Tessone, F Schweitzer: Betweenness preference: Quantifying correlations in the topological dynamics of temporal networks, Phys Rev Lett, 110(19), 198701, May 2013
This hands-on tutorial introduces the theoretical foundations of pathpy
and uses empirical and synthetic data to show how they can be practically applied in python
. The latest release of pathpy
is available via the python package index. In python 3.x, you can install it by typing:
pip install pathpy2
pathpy
is fully integrated with jupyter
, providing in-line, interactive and dynamic visualisations of graphs, temporal networks, as well as higher- and multi-order network models. This teaser highlights some of its features:
A description of the recommended setup to complete this tutorial is available online. A brief introduction to interactive data science with python
, Visual Studio Code
, and jupyter
is given in unit 1.
The remaining seven units (approx. 30 minutes each) introduce pathpy
’s approach to the modeling and analysis of time series data on complex networks. For each unit we provide a stand-alone HTML file, as well as a juypter notebook that you can download and run on your machine. In units 5 and 8 we invite you to use pathpy
to explore higher- and multi-order models on your own.
pathpy
is brought to you by the Data Analytics Group at the Department of Informatics of University of Zurich.
Feel free to get in touch if you want to host an interaction hands-on tutorial in your research group, institute, or company.