No, LOAD statements (where you will use your aggregation functions) are working on a single input table.
There is no logical inference engine during script time, compared to your data model after the LOAD script has finished.
So you would need to use joins / mappings etc. to get all fields needed into the table where you want to calculate the measures.
To be more precise, there are few methods to access field values from a different resident table, like
But the performance will be notable worse compared to mapping the field values to the same table, except for very simple settings.