brms with broom.mixed for models without group effects

[corriebar]

Hey,

I've been trying to follow a blog post on using brooms with brms. After failing to make it work with just brooms and finding out it moved to broom.mixed, I now get the error that my model doesn't have any group effects. Now, if I understand correctly, this package is supposed to be for mixed/hierarchical models. Is it still planned to be able to use broom with brms also for non-hierarchical models?

library(tidyverse)
library(brms)
library(broom.mixed)

set.seed(1)

n <- 50
mu_c <- 0
mu_t <- 0.5
d <-
  tibble(group     = rep(c("control", "treatment"), each = n)) %>% 
  mutate(treatment = ifelse(group == "control", 0, 1),
         y         = ifelse(group == "control", 
                            rnorm(n, mean = mu_c, sd = 1),
                            rnorm(n, mean = mu_t, sd = 1)))

fit <-
  brm(data = d,
      family = gaussian,
      y ~ 0 + Intercept + treatment,
      prior = c(prior(normal(0, 2), class = b),
                prior(student_t(3, 1, 1), class = sigma)),
      seed = 1)

tidy(fit)
#> Error: The model does not contain group-level effects.

Created on 2020-08-27 by the reprex package (v0.3.0)

[bbolker]

This seems to be a relatively minor tweak. I believe your example should work now if you reinstall (remotes::install_github("bbolker/broom.mixed")) ... let me know?

[corriebar]

It first showed the same error after reinstalling but worked after restarting the session.