I also can't get the confidence intervals when using a gamlss formula that has a smoother term in it. I don't think this is exactly the same problem but I just thought it was a good idea to add it here.
For example
gamlss_mod <- gamlss(x ~ y + ba( ~ s(z, by = w)))
ggpredict(gamlss_mod, terms = c("z", "y"))
new prediction
Error: Confidence intervals could not be computed.
* Reason: "object 'z' not found"
* Source: eval(predvars, data, env)
I also can't get the confidence intervals when using a gamlss formula that has a smoother term in it. I don't think this is exactly the same problem but I just thought it was a good idea to add it here.
For example
Originally posted by @Ari04T in https://github.com/strengejacke/ggeffects/issues/159#issuecomment-812148583