I’m just happy to be useful. If the GP Stan code is too much I think BRMS has GP regression now.
Yeah I think the the former is straightforward to model. The later part (time dependency of canopy) I have no idea, but I’m happy to look into it.
Thanks, I was going to ask. Can you please just share the plain csv/txt file, not the R data file, with out all the “shard” stuff? I can’t promise anything immediately, because I have projects due, but I’m happy to help.
I’ll post all the code/plots of everything I produce, so that way you have the resources to produce it on your own.
Also might be useful is the document I wrote up in the first post here: Applied Gaussian Processes in Stan, Part 1. A Case Study