The electrostatic graph algorithm: a physics-defined method for converting a time-series into a weighted complex network

TL;DR

This paper introduces a physics-inspired electrostatic graph algorithm that converts time-series data into weighted complex networks, providing a more structurally relevant representation compared to existing methods like the visibility graph.

Contribution

The paper presents a novel electrostatic-based method for transforming time-series into weighted graphs, enhancing structural relevance and offering a new framework for analysis.

Findings

Electrostatic graphs better capture the structure of time-series.

The method produces weighted, sometimes disconnected graphs.

It outperforms the visibility graph in structural relevance.

Abstract

This paper proposes a new method for converting a time-series into a weighted graph (complex network), which builds on the electrostatic conceptualization originating from physics. The proposed method conceptualizes a time-series as a series of stationary, electrically charged particles, on which Coulomb-like forces can be computed. This allows generating electrostatic-like graphs associated to time-series that, additionally to the existing transformations, can be also weighted and sometimes disconnected. Within this context, the paper examines the structural relevance between five different types of time-series and their associated graphs generated by the proposed algorithm and the visibility graph, which is currently the most established algorithm in the literature. The analysis compares the source time-series with the network-based node-series generated by network measures that are arranged into the node-ordering of the source time-series, in terms of linearity, chaotic behaviour, stationarity, periodicity, and cyclical structure. It is shown that the proposed electrostatic graph algorithm produces graphs that are more relevant to the structure of the source time-series by introducing a transformation that converts the time-series to graphs. This is more natural rather than algebraic, in comparison with existing physics-defined methods. The overall approach also suggests a methodological framework for evaluating the structural relevance between the source time-series and their associated graphs produced by any possible transformation.