I’m trying to run a count model using a population offset term. It’s going to end up being a zip, nb, or hurdle model in the end, but whenever I include population as an offset it fails with the following error. I don’t get this error when using population as a linear predictor – only as an offset.
mod1 <- brm( cases ~ offset( Pop ) , family=“zero_inflated_poisson” , data = bg)
SAMPLING FOR MODEL ‘e91f6f9e27134c8641c6a1986e72ed9b’ NOW (CHAIN 1).
Chain 1: Rejecting initial value:
Chain 1: Log probability evaluates to log(0), i.e. negative infinity.
Chain 1: Stan can’t start sampling from this initial value.
Chain 1: Rejecting initial value:
Chain 1: Log probability evaluates to log(0), i.e. negative infinity.
etc
I probably need some help specifying initial values if that’s the issue (I don’t seem to be doing it correctly) - but is that the problem? That I need to supply a list of 4 starting values (one for each chain) for the Pop parameter?