What is the SQL statement that removes duplicates but keep additional column's data?

I'd like to know what should be the SQL statement (for ORACLE DBMS) that would get back unique (by CUSTOMER_ID, VEHICLE_ID, DEA开发者_如何学CLER_ID and EVENT_TYPE_ID) rows BUT it will return the latest date (EVENT_INITIATED_DATE) for that row too. I've tried DISTINCT and GROUP BY, but wasn't able to figure out how to handle EVENT_INITIATED_DATE (that is DATE data type).

CUSTOMER_ID            VEHICLE_ID             DEALER_ID  EVENT_TYPE_ID          EVENT_INITIATED_DATE      
---------------------- ---------------------- ---------- ---------------------- ------------------------- 
22197630               23093399               6040       20                     11-JAN-07                 
22197630               23093399               6040       5                      11-JAN-07                 
22197630               23093399               6040       4                      11-JAN-07                 
22197630               23093399               6040       3                      11-JAN-07                 
22197630               23093399               6040       4                      19-JAN-06                 

select CUSTOMER_ID, VEHICLE_ID, DEALER_ID, EVENT_TYPE_ID, 
    max(EVENT_INITIATED_DATE)
from MyTable 
group by CUSTOMER_ID, VEHICLE_ID, DEALER_ID, EVENT_TYPE_ID

Distinc won't work.

Group by - and then one of the MAX / MIN etc. functions for the additional fields.

SELECT Customer_ID, VEHICLE_ID, MAX(EVENT_INITIATED_DATE)... ...GROUP BY Customer_ID, VEHICLE_ID

Hope that helps ;)

select CUSTOMER_ID, VEHICLE_ID, DEALER_ID, EVENT_TYPE_ID,
EVENT_INITIATED_DATE
from MyTable a 
group by CUSTOMER_ID, VEHICLE_ID, DEALER_ID, EVENT_TYPE_ID,
EVENT_INITIATED_DATE HAVING EVENT_INITIATED_DATE = 
  ( select MAX(EVENT_INITIATED_DATE) from MyTable WHERE CUSTOMER_ID =
     a.CUSTOMER_ID AND VEHICLE_ID = a.VEHICLE_ID AND DEALER_ID =
     a.DEALER_ID AND a.EVENT_TYPE_ID = EVENT_TYPE_ID );