Hello! I am new to coding these kinds of models and I was wondering if doing a changepoint model with brms with these data would be appropriate.
Here’s the gist: There are a bunch of tadpoles that go through metamorphosis (stages) but their growth rate throughout development isn’t consistent. I was hoping to use a changepoint model to show the thresholds where the rate of change between metamorphic time points changes.
I keep seeing examples of people using “step” but I am not sure what my “omega” would be in this case (referring to this case here from @ paul.buerkner, Piecewise Linear Mixed Models With a Random Change Point)
mod_1<-
brms::brm(
Weight ~ Treatment * step(Stage - ??? ) + (1 | ID),
family = student(link = “identity”),
data = growthrate,
cores = 4,
warmup = 50000,
thin = 2,
iter=100000))
Any insight or suggestions for how to approach this from a bayesian perspective would be great!
Chloe