I want to do a query that returns the employees who have no rows in the employeeunitJobtype table, so I do this query, and just for testing I add the employeeID of this one employee that I know doesn't have a row in the table:
SELECT * FROM Employee WHERE Employee.EmployeeId NOT IN (select EmployeeId from EmployeeUnitJobType) AND EmployeeId = 159795
Both work, but I do prefer the second, although it might not be as politically correct. We have a lot of units and jobs per employee and I'm thinking the second might return multiple rows and I only want one row returned.
thanks!
Scott Pletcher
"NOT IN" is roughly equivalent to a "NOT EXISTS", both of which take advantage of the anti semi join. SQL won't usually convert a left join and check for NULL into an anti join, and, if not, it's usually less efficient.
Btw, DISTINCT performs better than GROUPing when all you need to do is get distinct values, so if you do decide to use that approach, code it like this instead:
LEFT JOIN (SELECT DISTINCT EmployeeId from EmployeeUnitJobType) t