English

Pair circulas modelling for multivariate circular time series

Methodology 2023-11-23 v1

Abstract

Modelling multivariate circular time series is considered. The cross-sectional and serial dependence is described by circulas, which are analogs of copulas for circular distributions. In order to obtain a simple expression of the dependence structure, we decompose a multivariate circula density to a product of several pair circula densities. Moreover, to reduce the number of pair circula densities, we consider strictly stationary multi-order Markov processes. The real data analysis, in which the proposed model is fitted to multivariate time series wind direction data is also given.

Keywords

Cite

@article{arxiv.2311.13131,
  title  = {Pair circulas modelling for multivariate circular time series},
  author = {Hiroaki Ogata},
  journal= {arXiv preprint arXiv:2311.13131},
  year   = {2023}
}
R2 v1 2026-06-28T13:28:10.205Z