UDF forces Select-SQL code to perform as SCAN, so writing SCAN loop could be the way to clarify the code with no loss of speed.

>Good point! I would not use UDF, but a derived table instead.
>
>>There are two problems with your second example:
>>1. You should specify field size when calling the function, e.g.
>>select 0000+GetExpendituresBySubelement(elementnum) as expended
>>2. Secondly, (more important) it is risky to change active work area within the function while running select-SQL, i.e. this code is not reliable.
>>
>>
>>>This function
>>>

>>>function GetExpendituresBySubElement
>>>lparameters lnSubElement
>>>local lnOldWorkArea, lnRetVal
>>>dimension lnRetval(1)
>>>lnOldWorkArea = select(0)
>>>
>>>select  sum(amount) ;
>>>from tip ;
>>>inner join ledger on ledger.tip_id = tip.tip_id and source="RTA" ;
>>>where val(substr(sponsor_id,5,2)) = lnSubelement;
>>>into Array lnRetVal
>>>
>>>select(lnOldWorkArea)
>>>return iif(isnull(lnRetVal),0,lnRetVal)
>>>

>>>seems to work OK when called with a single value like this
>>>

>>>? GetExpendituresBySubelement(2)
>>>

>>>But when embedded in another select like this
>>>

>>>select *, GetExpendituresBySubelement(elementnum) as expended ;
>>>from rta_elements ;
>>>order by elementnum
>>>

>>>I get only 0's and *'s. Can someone please explain?
>>>
>>>Thanks