Tests Based on Ordinal Patterns

Ordinal patterns describe the dynamics of a time series by looking at the ranks of subsequent observations. By comparing ordinal patterns of two times series, Schnurr (2014) defines a robust and non-parametric dependence measure: the ordinal pattern coefficient. Functions to calculate this and a method to detect a change in the pattern coefficient proposed in Schnurr and Dehling (2017) are provided. Furthermore, the package contains a function for calculating the ordinal pattern frequencies. Generalized ordinal patterns as proposed by Schnurr and Fischer (2022) are also considered.