Skip to content

Predicting graph structure from a time series of graphs

This is a Python implementation of netseer.
Netseer is a tool that outputs a predicted graph based on a time series graph sequence

Purpose

The goal of netseer is to predict the graph structure including new nodes and edges from a time series of graphs.
The methodology is explained in the preprint (Kandanaarachchi et al. 2025).

Image of a time-series list of graphs and a predicted graph.

Installation

This package is available on PyPI, and can be installed with PIP or with a Package Manager:

pip install netseer # or uv add netseer

Quick Example

Generating an example graph list:

from netseer import generate_graph

graph_list = generate_graph.generate_graph_list()

The generate_graph_list() function has parameters for templating what types of graphs to generate. Information about these can be found in the reference docs.

Predicting on that graph:

from netseer import prediction

predict = prediction.predict_graph(graph_list, h=1)

Increasing the h parameter increases how many steps into the future the prediction is, with h=1 being 1 step in the graph sequence.

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.