Hi,

I would like to simulate a response using `posterior_predict()`

with a model where `sample_prior='only'`

was specified. As written in the docs, this gives me a matrix with `n_samples`

rows and `N`

columns.

- Is it possible to somehow also know which exact samples from the prior were used in order to simulate each row of size N?
- How does
`posterior_predict`

pick a concrete sample from the prior distributions, randomly or just picking the already sampled values from the prior one by one? -
`sample_prior='only'`

does not store the priors, is this intended?

Thanks in advance for your help!

R version: 4.0.3

brms: 2.14.4