Does anyone know the rule of them [Edit: thumb?] about using aggregation at the object level?
Is it better to the sum function around the object?
i.e. SUM(table.colom) therefore creating a group by clause in the SQL
or
Is it better to set the property function to sum in designer?
Last questions if using the SUM(table.colom) in the object SQL what do you set the aggregation property to?
Any help on this topic would be greatly appreaciated.
The answer is both. Putting sum() in the select does indeed generate a group by for all of the non-aggregated columns in a query. Always best to let the database do the work.
The aggregation property controls similar behavior, but for the reporting tool, not SQL. In other words, if you remove a dimension from a report block (table), the aggregation property will cause the measure to aggregate automatically.
(BOB member since 2003-02-13)
(BOB member since 2002-06-20)