question

Still indirectly, yes, with more labels. But you are responsible to allow only values that makes senses. The idea is to evaluate a distribution as a number by a switch by case structure, which matches with your input field. Each statistical distribution parameter is a label value.

An example:

A label value of 3 is “normal” for ”labeldistro”, then “labelparam1” is mean and “labelparam2” is deviation. Any other labels aren’t of any use.

switch(current.labeldistro){
   case 3:
     return normal(current.labelparam1,current.labelparam2,getstream(current));
…}

Edit: I replaced break interrupt by return structural command