Dear @AnestisTouloumis ,
Would it be very complicated technically to make the solvers (for example ordLORgee) work with multiply imputed datasets of class "mids"? This is an object created by the mice package, storing all multiply imputed datasets.
Unfortunately, when I try to fit the model through the mice approach using its "with()" function, it fails:
with(imp_data_longit,
ordLORgee(formula = ODIPain ~ Visit * Arm + Age_centered + Age_centered : Visit,
id = PatientId,
repeated = as.numeric(Visit),
LORstr = "category.exch",
link = "logit")
)
Error in eval(predvars, data, env) : object 'ODIPain' not found
while this column does exist in the data:
I would like to use it with the emmeans package, which can handle a list of multiply imputed models, but this way I'm unable to.
Dear @AnestisTouloumis , Would it be very complicated technically to make the solvers (for example ordLORgee) work with multiply imputed datasets of class "mids"? This is an object created by the mice package, storing all multiply imputed datasets.
Unfortunately, when I try to fit the model through the mice approach using its "with()" function, it fails:
while this column does exist in the data:
I would like to use it with the emmeans package, which can handle a list of multiply imputed models, but this way I'm unable to.