drop levels in tableby() in grouped data

[kekec3778]

Is it possible to manage drop levels in tableby() function in grouped data? I read that unused levels are automatically dropped. Maybe integrate some argument if I want to.
And another issue I think is somehow connected with this, that when I have variable with only missing data, tableby() will not show anything, just Warning message Zero-length levels found for ...... Maybe will be better if on output is variable with information of the count of missing (Nmiss).