I need to give students a practical session on solving runtime and convergence issues using simulated data. However, how to simulate these warnings one by one?
Data: 100 individuals, 70% female and they are 10 years younger than males (this can be changed however needed)
data = bind_rows( tibble(sex = rep("Female", length.out = T), age = rnorm(70, 40, 4)), tibble(sex = rep("Male", length.out = T), age = rnorm(30, 50, 4))) m = brm(age ~ sex, data = data) m
How should I edit the simulated data and/or model so that it starts showing
- high Rhat values or
- low Bulk_ESS or
- low Tail_ESS or
- divergent transitions?
Your help would be much appreciated!