Is there a way to avoid the for loop? With my previous installation the for loop was running (it was very slow but produced properly), but since i installed bambi it must have updated some libraries and it runs only for a small number (<20). That’s why i asked if there is a more proper way.
Thanks again.
Yes, with some slighly complex advanced indexing.
Are the average running times for the same runner across multiple years related in the prior? If not, you can treat each runner and year combination as it’s own identity: have one parameter for each, ravel your observations and use a single vector advanced indexing to select the right means for each observation.
If they are related, it’s more tricky because you will have two levels of prior. Before we go there, just confirm if that’s the case.