I am learning and testing data modules as of recently. Whenever I add a new table from my source (SQL Server database), the automatic classification of my columns is generally good except that it always sets the Aggregation to COUNT DISTINCT for Attributes and Identifiers. 

From what I read in IBM documentation it selects an aggregation based on the source... Is there any way to change the default to be NONE?

Thank you

------------------------------
Guy-Philippe Arseneault
------------------------------
#CognosAnalyticswithWatson

I'm assuming that you are referring to attributes. 

Select the query item and choose properties from the context menu.

Click on the aggregate drop down.  Scroll to the top.  The top item in the list is none.  

Just out of curiosity, is there any particular reason you object to the default aggregation being count distinct?

------------------------------
IAN HENDERSON
------------------------------

Thanks for the reply.

It might be down to my lack of experience with the Data modules, but I thought that count distinct would make it so my attribute would aggregate as count distinct in reports by default and not do a group by. But instead it seems to be defaulting to no aggregation in the report. 

I was not aware of this behavior, which means I probably don't have to make them all aggregate to none? 






------------------------------
Guy-Philippe Arseneault
------------------------------

Original Message:
Sent: Mon August 24, 2020 03:33 PM
From: IAN HENDERSON
Subject: Default Aggregate Type in Data Module

I'm assuming that you are referring to attributes. 

Select the query item and choose properties from the context menu.

Click on the aggregate drop down.  Scroll to the top.  The top item in the list is none.  

Just out of curiosity, is there any particular reason you object to the default aggregation being count distinct?

------------------------------
IAN HENDERSON
------------------------------

That's largely determined by the usage.

In addition, if you have column dependency, you have some further control over the SQL generation of the attribute. If the setting is minimum or maximum, the attribute will be either MINed or MAXed in the select clause. If it is group by, the attribute will be added to the group by clause.

This is shown in the illustration below.  On the left the SQL generation setting of the product name attribute has been set to group by (from the default of minimum).  On the right it has been set to maximum.

------------------------------
IAN HENDERSON
------------------------------

Original Message:
Sent: Tue August 25, 2020 09:36 AM
From: Guy-Philippe Arseneault
Subject: Default Aggregate Type in Data Module

Thanks for the reply.

It might be down to my lack of experience with the Data modules, but I thought that count distinct would make it so my attribute would aggregate as count distinct in reports by default and not do a group by. But instead it seems to be defaulting to no aggregation in the report. 

I was not aware of this behavior, which means I probably don't have to make them all aggregate to none? 






------------------------------
Guy-Philippe Arseneault
------------------------------


Original Message:
Sent: Mon August 24, 2020 03:33 PM
From: IAN HENDERSON
Subject: Default Aggregate Type in Data Module

I'm assuming that you are referring to attributes. 

Select the query item and choose properties from the context menu.

Click on the aggregate drop down.  Scroll to the top.  The top item in the list is none.  

Just out of curiosity, is there any particular reason you object to the default aggregation being count distinct?

------------------------------
IAN HENDERSON