Yes — you would have to adjust the script to handle more segments but the Stan model should work fine. No haven’t done any benchmarks
Adjusted, tested with my not-so-small dataset and it’s awfully slow. 2 hours for 100-200 samples, then I stopped it. Do you have some ideas on how to speed it up?
I’ve read somewhere that STAN is generally relatively slow due to the NUTS sampler, compared to e.g. Gibbs.