Complex hierarchy in bambi, is this possible?

Questions version agnostic
1

I’m using a GLM to model some treatment effects, in the simplest form it looks like this:

mu = 1 + treatment:individual

however my subjects are categorized in groups, I want to build a hierarchy, so that the individuals within the group pool the data across each other, but not between the groups. I tried this:

mu = 1 + treatment:individual|group

but it does not work as expected.

1 Like
2

CC @tcapretto

3

Could you tell me a bit more about the problem and the data? Are subjects exposed to multiple treatments? Or why are you looking for the interaction between treatment and individual?