Question about format for sending data to Stan for compartmental PK models

Modeling
1

I’m a noob and still writing the Stan program.

In the PK compartmental model, I have a parent compound with three metabolites. Hence, I’m expressing the amounts of material in the compartments in micromoles to save having to account for differences in molecular weights.

When I supply the data to stan, would it be best to put it into units of micromoles, log(micromoles) or something else? The data was presented in the original paper as ng/ml.

Thanks for any help in advance.

2

Scale things so that parameters are of the order of unity. That works best. Often this means to deal with logs.