A 'group by' column is needed when performing an aggregation.
This can be:
- during an aggregation within a table, or
- during a join, when the values that you wish to joined need to be aggregated; the selected key 2 column will act as a 'group by' column.
When a 'group by' column is used:
- groups are formed for all records in the table that have the same class name, or ID in the specified 'group by' column,
- then, the specified aggregation function calculates the sum, average, etc. of values per group.
See also:
TabCalc : aggregate functions
TabCalc : joining columns from other tables