Question

I have a table like this:

id make model year
1 chevy camaro 1969
2 chevy camaro 1967
3 ford mustang 1967
4 chevy corvette 1984
5 pontiac gto 1971
6 pontiac grandam 1987

I want to return a recordset that includes one row (all columns) for each distinct value in make. Example of results:

1 chevy camaro 1969
3 ford mustang 1967
5 pontiac gto 1971

Order is not important to me; it does not need to be deterministic. This is so I can easily select out example data for different use cases without having to send the entire table contents to a developer.

Answer 1

In MySQL, you may be able to get away with:

SELECT id, make, model, year
  FROM AnonymousTableAboutCars
 GROUP BY make, model;

With most other SQL DBMS, you’d probably do:

SELECT id, make, model, year
  FROM AnonymousTableAboutCars AS a
  JOIN (SELECT MIN(id)
          FROM AnonymousTableAboutCars
         GROUP BY make, model
       ) AS m
    ON a.id = m.id;

The choice of aggregate is up to you; MAX springs to mind as a valid, deterministic alternative.