Null hypothesis testing: does x1 exclude n?

Modeling
2

I can use tidybayes::add_epred_draws to get draws from the expectation of the posterior predictive distribution to get the average expectation for each level (see here).

I can then use these to visualise expected distributions:


draws = iris %>% 
  distinct(Species) %>% 
  add_epred_draws(stan_lm1, scale = 'response', re_formula = NA, ndraws = 500)

draws %>% 
  ggplot(aes(x = Species, y = .epred)) +
  geom_violin() +
  geom_boxplot(width = .1)