Skip to contents

Predicting graph structure from a time series of graphs

If you have a time series of dynamic graphs with changing structure, how would you predict future graphs? This is the goal of netseer.

Netseer predicts the graph structure including new nodes and edges from a time series of graphs. It adapts Flux Balance Analysis, a method used in metabolic network reconstruction to predict the structue of future graphs. The methodology is explained in the preprint (Kandanaarachchi et al. 2025).

R package

The algorithm is available in both R and Python. The R package netseer in on CRAN and can be installed as follows:

install_packages("netseer")

The vignette for the R package is available under Get Started at https://sevvandi.github.io/netseer/

Python code

The Python code is available at https://github.com/sevvandi/netseer-python/. The Python package will be made available in the near future.

Coding Credits

A big shout out to Stefan Westerlund and Brodie Oldfield for helping with this package. Stefan optimized the algorithm in R and coded it from scratch in Python. Brodie is currently packaging up the Python code and is working on a website for the Python version.

References

Kandanaarachchi, Sevvandi, Ziqi Xu, Stefan Westerlund, and Conrad Sanderson. 2025. “Predicting Graph Structure via Adapted Flux Balance Analysis.” https://arxiv.org/abs/2507.05806.