Level 200 - ETL
Accepted Answer
-
Hi @Walter ,
usually, the aggregation will take place in the mapping section of a datareader, using the ADD option for the cubes' slices. Consider this example:
This will sum up all values regarding the grouping by the cubes' dimensions (Month, Store, Currency, and Account). Anything configured in the ETL will work on row level (i.e. being calculated per 1 row at the time).
Another possibility of aggregating would be to use a cube that is higher aggregated regarding its dimensions than the source data is. If, for example, the above used cube was dimensioned by quarter instead of month, it would automatically aggregate all month's values into the corresponding quarter.
What do you mean by "aggregation between two or more columns"? The columns in the ETL can be referred to the same way you'd do in EXCEL (i.e. D7=C7*E7 would multiply C7 by 2 and store the result in column D7).
Best,
Helmut1