Use formula in `brms` format i.e. `brms::bf` as input to `epidist`

epinowcast / epidist

An R package for estimating epidemiological delay distributions

http://epidist.epinowcast.org/

Other

12 stars 4 forks source link

Use formula in `brms` format i.e. `brms::bf` as input to `epidist` #195

Closed athowes closed 2 months ago

athowes commented 2 months ago

And should this instead be in the brms format? i.e. brms::bf or something like that. Rather than a list of formula

_Originally posted by @athowes in https://github.com/epinowcast/epidist/pull/183#discussion_r1689793381_

We should try to use the same input methods for formula as brms does. This would also mean that users wouldn't have to input formula for all variables, and it would know that if a formula is not provided then to use not_provided ~ 1.

athowes commented 2 months ago

brms uses an argument formula and then passes additional formula via ...
There is a function brms:::validate_formula that can take at least

formula <- validate_formula(formula, data = data, family = family, autocor = autocor, sparse = sparse, cov_ranef = cov_ranef)

https://github.com/paul-buerkner/brms/blob/2917e72ec0cbd9cd8ade6465cf44cfaf7c457461/R/brmsformula.R#L1249

athowes commented 2 months ago

Closing this allows users to fix distributional parameters: