Orlace删除重复行

SQL> select deptno,dname,loc,rowid from mydept order by deptno;

    DEPTNO DNAME          LOC           ROWID
---------- -------------- ------------- ------------------
        10 ACCOUNTING     NEW YORK      AAAM6GAAEAAAAG8AAA
        11 ACCOUNTING     NEW YORK      AAAM6GAAEAAAAG9AAA
        12 ACCOUNTING     NEW YORK      AAAM6GAAEAAAAG9AAB
        13 ACCOUNTING     NEW YORK      AAAM6GAAEAAAAG9AAC
        20 RESEARCH       DALLAS        AAAM6GAAEAAAAG8AAB
        21 RESEARCH       DALLAS        AAAM6GAAEAAAAG9AAD
        30 SALES          CHICAGO       AAAM6GAAEAAAAG8AAC
        40 OPERATIONS     BOSTON        AAAM6GAAEAAAAG8AAD

DELETE FROM mydept

WHERE ROWID NOT IN (

   SELECT MIN(ROWID) FROM mydept

   GROUP BY dname,loc);