"unexpected end of input" error when running loo on a brms fit

[bnicenboim]

I'm running the following model

m <-  brm(bf(n400 ~ c_cloze 
                                     + (1 | item) + (c_cloze  | subject)+
                                     + (c_cloze  | lab), sigma ~ (1|lab)),
                  prior= c(
                      prior(normal(0, 10), class = Intercept),
                      prior(normal(0, 1), class = b),
                      prior(normal(0, 10), class = sd),
                      prior(lkj(2), class = cor),
                      prior(normal(0,10), class = Intercept, dpar= sigma),
                      prior(normal(0,10), class = sd, dpar= sigma)
                  ),
                  data = data_eeg_noun_nieuwland,
                  chains = chains, iter=iter,
                  control= list(adapt_delta = .99, max_treedepth=13)
)

Then when I do

 loo_raw <- loo(m)

I get the following error:

Error in parse(text = x, keep.source = FALSE) : 
  <text>:2:0: unexpected end of input
1: n400~c_cloze+
   ^

You can find a subset of the data (that give the same problem) here: www.ling.uni-potsdam.de/~nicenboim/files/data_loo.RDS

If I remove some random effects, it actually works, so I'm not sure of what's going on.

[paul-buerkner]

There is a typo in your formula having ++ in it after the varying effect of subject. This should ideally not create any problems, but apparently it does. I will see how I can fix it, but in the mean time you can just remove the extra + and that's it.

[paul-buerkner]

should be fixed now.