I am looking into your stanfile, and if learning stan syntax for hierarchical dynamic model such as ragged array is the goal, would you recommend starting from this ragged.stan file? I see you have created many stan files and your insight on which syntax would be most general (e.g. with more complex hierarchy structure) and efficient would be helpful.
Also, another question is dt in datablock from this line of yours. Some dynamic modelers call this Time_step and perceive it as a hidden variable that shouldn’t affect the model. They decrease by half until it doesn’t affect state time-series too much. However, from @andrewgelman’s The Pinocchio Principle: A model that is created solely for computational reasons can take on a life of its own, receiving time_step in data block (as here) was my first cut which couldn’t get much support. How did you set dt and is this relevant to precision?
tagging @OriolAbril and @ahartikainen who I’m collaborating with for transplanting the above output to InferenceData with (prior_draws, chain, draws, hierarchy_idx) coordinate.