Unique values table

A unique values table (UVT) contains all unique values that exist in the source data for a specific configuration variable. When displaying all values for a variable, the Oracle Empirica Signal application references the unique values table for that variable. (The exception to this is on the Select Criteria page for results of a data mining run, where users select from a list of values in the data mining results.)

When you define data configuration variables, if a variable's selection type is Distinct Value List, Select from Drug Hierarchy, or Select from Event Hierarchy, you can also specify a unique values table. If you do not specify a UVT for a variable, the application creates the corresponding UVT automatically when you validate the data configuration. If the source data is updated, as in a timestamped data configuration, you must also update the unique values tables or delete them for the individual variables so that they will be re-created (and updated) automatically during validation.

A unique values table name starts with UVT by default. The table must have an ITEM_VALUE column, which contains one row for each unique value. The column type of the ITEM_VALUE column must match the column type of the corresponding column in the source database table.

If text values for a variable are mapped, the unique values table contains source data values, not the mapped values.

Note:

In a timestamped database there are no start dates and end dates associated with the unique values table, so all unique values that ever existed are in the unique values tables.

A superuser can view the unique values tables on the Data Sources page.