Primary Key Discovery
To discover possible keys in source tables, create a primary key profile.
To identify potential data quality issues, create a primary key profile that discovers non-unique rows. A primary key discovery profile is also useful when the primary key is a combination of columns that uniquely defines a row. Non-conforming rows might contain duplicate information. After you discover potential keys in a table, you can add the keys to the source table in a project.
For example, an Employee table contains an EmpID column. The value of the column is 100% unique. A primary key profile identifies the column as a possible primary key that you can add to the table in the project.
Primary Key Profile Options
When you run a primary key profile, the Data Integration Service infers primary key candidates from the source tables. To limit the candidate results, configure inference options.
A primary key profile identifies the columns and combinations of columns that meet a specific conformance level. You can edit the maximum number of columns to combine for unique key identification.
The following table describes the options for primary key profiles:
Option | Description |
|---|
Primary Key Rows | The maximum number of columns that the profile can combine to form a primary key. Default is 3. |
Max Rows | The maximum number of rows to profile. Default is 1,000. |
Conformance Criteria | The minimum percentage or maximum number of key violations that the profile allows. Default is a minimum percentage of 99. |