[R-sig-ME] Likelihood drops on adding random effect

[Ben Bolker]

Murray Jorgensen <maj at ...> writes:

> 
> *I tried using 'update' which should (I believe) initialise at the simpler
> model:
> 
> > f0128biu = update(f0128bi, . ~ . + (1|tree))
> Error in model.frame.default(formula = fincr ~ icfac + (1 + gap) + (1 +  :
>   invalid type (closure) for variable 'box'
> 
> Murray
> 

  It's not really shocking to me that update() doesn't work, fiddling
with random effects formulas can be tricky.  Are you willing to post
data somewhere (or e-mail it to me)?

  Ben Bolker