# Ordinal Poincar\'e Sections: Reconstructing the First Return Map from an Ordinal Segmentation of Time Series

## TL;DR

This paper introduces a new method for reconstructing first return maps from time series data using ordinal partitions, eliminating the need for embedding and improving robustness and efficiency in analyzing chaotic systems.

## Contribution

The authors present a novel ordinal-based algorithm for constructing first return maps directly from time series without embedding, enhancing robustness and computational efficiency.

## Key findings

- Successfully applied to Lorenz, Rössler, and Mackey-Glass systems
- Provides a robust alternative to traditional geometric methods
- Guided by entropy measures for optimal ordinal sequence selection

## Abstract

We propose a robust and computationally efficient algorithm to generically construct first return maps of dynamical systems from time series without the need for embedding. Typically, a first return map is constructed using a heuristic convenience (maxima or zero-crossings of the time series, for example) or a computationally delicate geometric approach (explicitly constructing a Poincar\'e section from a hyper-surface normal to the flow and then interpolating to determine intersections with trajectories). Our approach relies on ordinal partitions of the time series and builds the first return map from successive intersections with particular ordinal sequences. Generically, we can obtain distinct first return maps for each ordinal sequence. We define entropy-based measures to guide our selection of the ordinal sequence for a ``good'' first return map and show that this method can robustly be applied to time series from classical chaotic systems to extract the underlying first return map dynamics. The results are shown on several well-known dynamical systems (Lorenz, R{\"o}ssler and Mackey-Glass in chaotic regimes).