Tensor-product interactions in Markov-switching models

This repository provides code to reproduce the case studies presented in the paper “Tensor-product interactions in Markov-switching models”.

In the paper I discuss incorporating tensor-product interactions into Markov-switching models to capture complex temporal and spatial patterns in animal movement and behaviour data. The approach allows for flexible modelling of time-varying effects, enabling the analysis of intricate relationships between multiple covariates over time and space. Efficient anisotropic smoothness selection is conducted approximately using the so-called extended Fellner-Schall method.

The methodology is illustrated through three applied case studies analysing

1. diel variation of an African elephant across the annual cycle,
2. genotype-specific behavioural rhythms in fruit flies, and
3. time-varying spatial effects in Arctic muskox movement data.

Code for reproducing all results can be found in the case_studies directory.