Identifying Out-of-Date or Obsolete Values in Business Data
To find values that are no longer valid in a data set, configure a rule statement to read a reference table that contains known data errors. For example, a national organization updated its branch codes from an eight-digit format to a ten-digit format. You might configure a rule statement to read a reference table that contains the obsolete eight-digit codes.
If the rule statement finds an obsolete branch code in the input data, it writes a value that you specify as output. If the rule statement does not find an obsolete branch code, the action writes the input value as output. The action writes an output for each input row that satisfies the condition in the rule statement.
1. Select a rule set in a rule specification.
2. In the rule set properties, click Rule Logic.
3. Click Add Rule Statement.
4. Select an input for the rule statement.
- - If the rule set does not contain an input, create an input. Configure the input properties to represent the branch code data.
5. Select an operator to validate the results of the condition analysis.
- - To specify that the input data must match the data in the condition, select the following operator:
is
6. Select the type of condition to apply to the input.
- - To compare the input data to the data from another input, select the following condition type:
input
7. Select the input that you selected in step 4.
The rule statement compares the input to itself. Therefore, the rule statement passes all of the input data values from the condition to the action.
8. Configure the action to search the input data for obsolete branch codes.
- - Open the Define Action dialog box.
- - Select Standardize, and select Replace Reference Table Matches with Custom Strings.
- - Select the input that you specified in the condition.
- - Select the reference table from the Model repository.
- - Enter the data value that the action returns when an input data value matches a reference table value.
9. Click OK.
10. Save the rule specification.