In SQL querying, which clause should be used to return only one copy of each set of duplicate rows from a result set (i.e., remove duplicates in the projection)?

Introduction / Context: When retrieving data, you may encounter duplicates due to joins, denormalization, or repeated values in a column set. SQL provides a standard way to collapse duplicates so that each unique combination of selected columns appears only once in the result set.

Given Data / Assumptions:

• We are writing a SELECT query that projects columns potentially containing duplicate combinations.
• We need the ANSI-standard clause that removes duplicates.
• Portability across common RDBMSs is desired.

Concept / Approach: The DISTINCT keyword is defined by the SQL standard to eliminate duplicate rows from the projection. Some vendors historically accepted UNIQUE as a synonym, but it is not universally portable. There is no standard DIFFERENT keyword for this purpose.

Step-by-Step Solution:

Verification / Alternative check: SQL references for major systems (PostgreSQL, SQL Server, MySQL, Oracle) document DISTINCT as the standard duplicate-elimination modifier in SELECT.

Why Other Options Are Wrong:

• SELECT UNIQUE: works in some dialects (e.g., older Oracle), not standard everywhere.
• SELECT DIFFERENT: not an SQL keyword.
• All of the above: incorrect because one option is non-standard and one is invalid.
• None: DISTINCT is the correct standard.

Common Pitfalls: Overusing DISTINCT to mask poor joins; DISTINCT adds sorting/aggregation overhead. Consider proper join conditions and indexes before relying on DISTINCT for performance-sensitive queries.

Final Answer: SELECT DISTINCT