To get the top-n row, use a WHERE clause in the statement.

For example, if we wanted to see the top five original salaries, we would use this query:


SQL> drop table emp;

Table dropped.-- from   www . ja  v  a2s .c o  m

SQL> create table emp(
  2  empno        Number(3)  NOT NULL,
  3  ename        VARCHAR2(10 BYTE),
  4  hireDate     DATE,
  5  orig_salary  Number(8,2),
  6  curr_salary  Number(8,2),
  7  region       VARCHAR2(1 BYTE)
  8  );
SQL> insert into emp values(1,'Alison',DATE '1996-03-21', 45000, NULL,  'E');
SQL> insert into emp values(2,'James',DATE  '1978-12-12', 23000, 32000, 'W');
SQL> insert into emp values(3,'Celia',DATE  '1982-10-24', NULL,  58000, 'E');
SQL> insert into emp values(4,'Robert',DATE '1984-01-15', 31000, NULL,  'W');
SQL> insert into emp values(5,'Linda',DATE  '1987-07-30', NULL,  53000, 'E');
SQL> insert into emp values(6,'David',DATE  '1990-12-31', 78000, NULL,  'W');
SQL> insert into emp values(7,'Jode',DATE  '1996-09-17', 21000, 29000, 'E');
  2  FROM
  3   (SELECT empno, ename, orig_salary,
  4     DENSE_RANK() OVER(ORDER BY orig_salary desc) toprank
  5  FROM emp)
  6  WHERE toprank <= 5;

We have to alias the rank and use the alias in the ORDER BY.

Related Topic