Building significant columns data profiling scripts

Significant columns are the proposed natural columns you will be using in your dimensions to uniquely identify an individual record for the dimension. These keys determine the grain of the dimension.

Getting ready

Identify all the dimension entities within your semantic data model and determine the source tables for these dimensions.

How to do it...

Significant columns may differ from the natural key of the table:

1. 1. Connect to the source system using Oracle SQL Developer.

2. 2. Build a SQL statement to validate the grain of the dimension.

   Sample SQL statement:

   select <attribute_name>, <attribute_name>,
   <attribute_name> , count(*)
   from <schema.table_name>
   group by <attribute_name>, <attribute_name>,
   <attribute_name>;
   

   Sample SQL statement:

   select planner_code, segment1 , count(*)
   from inv.mtl_system_items
   group by planner_code, segment1;
   

3. 3. Validate the result set.

How it works...

The results which you...