mice with() function

[nsdelablancac]

Hi all!
I'm using mice package following recommendations of Zhang here in this paper (DOI: 10.3978/j.issn.2305-5839.2015.12.63). After completing mice() step, with() function does not work with my mids output object:

  #mice version: 3.8.0 2020-02-12
  with_data <- with(data = imp,
                       expr = glm(sex ~ variable1+variable2+variable3+variable4...))
  Error in `contrasts<-`(`*tmp*`, value = contr.funs[1 + isOF[nn]]) : 
  contrasts can be applied only to factors with 2 or more levels.

I checked my variable "sex" is as.factor.
with() function of mice() package does not appear in my R session, only with() from base, which I think it does not accept mids object and consequently it gives me that error. mice::with does not work neither.
When I ask for help (?with.mids) it appears, but documentation specify with() and with.mids() does not exists.
Does anyone know why it occurs? Thank a lot!

[gerkovink]

Your model has a predictor that is a factor, but has fewer than 2 levels. This is an error generated in glm and is not related to mice.

See e.g. the following reproducible example where the subset of Species results in a predictor with only one level.

lm(Sepal.Length ~ Sepal.Width + Species,
   data=iris[iris$Species == "setosa", ]) 

All the best,

Gerko

[gerkovink]

Closing as it is not related to mice.

[gerkovink]

By the way: with.mids() is not an exported object, but can be directly accessed with mice:::with.mids()

[nsdelablancac]

Thanks a lot! I could solve the problem.