In addition to supporting brms' existing model-based imputation feature #43 , I would like to propose adding support for multiple responses #42 where a strict subset of responses are fully unobserved. A trivial case of this is specifying random effects in separate formulas:
xf <- bf(x_latent ~ 1)
yf <- bf(y ~ x_latent)
Semantically, this would be identical to the partially missing data case; it would merely be a syntax extension to add additional column(s) for the fully unobserved variable(s) to a user's input dataframe.
Extends #43.
In addition to supporting brms' existing model-based imputation feature #43 , I would like to propose adding support for multiple responses #42 where a strict subset of responses are fully unobserved. A trivial case of this is specifying random effects in separate formulas:
Semantically, this would be identical to the partially missing data case; it would merely be a syntax extension to add additional column(s) for the fully unobserved variable(s) to a user's input dataframe.