If these are generics, then they could be extended to work with survey objects (currently, cardx::ard_svycategorical(). This could make downstream tabling would be easier. And we could also later extend it to data with multiple imputations.
I wonder if the processing function would benefit from the same treatment 🤔
If these are generics, then they could be extended to work with
survey
objects (currently,cardx::ard_svycategorical()
. This could make downstream tabling would be easier. And we could also later extend it to data with multiple imputations.I wonder if the processing function would benefit from the same treatment 🤔