RethinkDB: GROUP BY example

If you add a ‘group’ to a RethinkDB query, it will behave similarly to a similar query in SQL, except that you need to ‘ungroup’ it afterward:

r.db('test')
 .table('salaries')
 .group('role')
 .sum('salary')
 .ungroup()
 .map({Title: r.row('group'), Amount: r.row('reduction')})

If, however, you group by two items, the grouping will come out as an array, rather than individual, named columns:

r.db('test')
 .table('salaries')
 .group('role', 'level')
 .sum('salary')
 .ungroup()

7 rows returned in 61ms.

If this bothers you, and you want to make the result look like the output of SQL, you can add a mapping function to turn it back into rows, like so:

 .table('salaries')
 .group('role', 'level')
 .sum('salary')
 .ungroup()
 .map( 
   (x) => {
     return {
       Title: x('group')(0),
       Level: x('group')(1),
       Value: x('reduction')
     }
   } 
 )

7 rows returned in 54ms.