So I'm trying to compare a two-part model that incorporates random effects with a similar version that does not. I love how wonderful the mixed_model() function is, but I'm wondering if there's a way to run it without the random effect component.
The normal call (which uses a random effect by group) is below:
mixed_model(Y~ X,
random = ~ 1|group,
data=df,
family = hurdle.lognormal(), n_phis = 1,
zi_fixed = ~ Z")
However, whenever I set random = ~1 or random = NULL or random = ~0, I get an error. Is there any way to run this two-part hurdle model without using a random effect? I would really like to see how the model compares without it.
Unfortunately, this is not possible. The package is designed to work with random effects. There are other packages in R to fit this model without random effects.
So I'm trying to compare a two-part model that incorporates random effects with a similar version that does not. I love how wonderful the
mixed_model()
function is, but I'm wondering if there's a way to run it without the random effect component.The normal call (which uses a random effect by group) is below:
However, whenever I set
random = ~1
orrandom = NULL
orrandom = ~0
, I get an error. Is there any way to run this two-part hurdle model without using a random effect? I would really like to see how the model compares without it.Thank you so much for your time.