For example, if a Substitution rule was to be applied to the FIRST_NAME column of the EMPLOYEE table the relevance of the column could be preserved by substituting female and male names where appropriate. All female first names in the table could be substituted by choosing the Random Female First Names data set and using a WHERE Clause of WHERE EMP_GENDER='F'. This would cause only the female employee records to be selected and masked. A separate rule using the Random Male First Names data set and a second WHERE clause would have to be used to perform a similar substitution on the male entries.
The sampling option can also sift the set of rows to be operated on by the substitution rule. The sampling is done on whichever rows are returned by the WHERE Clause option. If no option is set, the sampling will apply to all rows in the table. If a WHERE Clause option has been configured, then the sampling will apply only to the rows returned by the WHERE Clause. For example, if the table contains 1000 rows and the sample percentage is sent to 10 percent then 100 rows (at random) will have the substitution operation applied to them. If a WHERE Clause was applied which reduced the selected rows from the full 1000 to 500, then the sample percentage of 10% would cause only 50 of the selected rows to have the substitution operation applied to them.