Mega Code Archive

 

Categories / Oracle PLSQL / Analytical Functions

 

If the statement were without NULLS LAST, the values of the NVLd nulls occur first

SQL> SQL> -- create demo table SQL> create table Employee(   2    empno              Number(3)  NOT NULL, -- Employee ID   3    ename              VARCHAR2(10 BYTE),   -- Employee Name   4    hireDate          DATE,                -- Date Employee Hired   5    orig_salary        Number(8,2),         -- Orignal Salary   6    curr_salary        Number(8,2),         -- Current Salary   7    region             VARCHAR2(1 BYTE)     -- Region where employeed   8  )   9  / Table created. SQL> SQL> SQL> -- prepare data for employee table SQL> insert into Employee(empno,  ename,  hireDate,                   orig_salary, curr_salary, region)   2                values(122,'Alison',to_date('19960321','YYYYMMDD'), 45000,       NULL,       'E')   3  / 1 row created. SQL> insert into Employee(empno,  ename,  hireDate,                       orig_salary, curr_salary, region)   2                values(123, 'James',to_date('19781212','YYYYMMDD'), 23000,       32000,       'W')   3  / 1 row created. SQL> insert into Employee(empno,  ename,  hireDate,                       orig_salary, curr_salary, region)   2                values(104,'Celia',to_date('19821024','YYYYMMDD'), NULL,       58000,        'E')   3  / 1 row created. SQL> insert into Employee(empno,  ename,  hireDate,                       orig_salary, curr_salary, region)   2                values(105,'Robert',to_date('19840115','YYYYMMDD'), 31000,      NULL,        'W')   3  / 1 row created. SQL> insert into Employee(empno,  ename,  hireDate,                       orig_salary, curr_salary, region)   2                values(116,'Linda', to_date('19870730','YYYYMMDD'), NULL,       53000,       'E')   3  / 1 row created. SQL> insert into Employee(empno,  ename,  hireDate,                       orig_salary, curr_salary, region)   2                values(117,'David', to_date('19901231','YYYYMMDD'), 78000,       NULL,       'W')   3  / 1 row created. SQL> insert into Employee(empno,  ename,  hireDate,                       orig_salary, curr_salary, region)   2                values(108,'Jode',  to_date('19960917','YYYYMMDD'), 21000,       29000,       'E')   3  / 1 row created. SQL> SQL> -- display data in the table SQL> select * from Employee   2  /      EMPNO ENAME      HIREDATE  ORIG_SALARY CURR_SALARY R ---------- ---------- --------- ----------- ----------- -        122 Alison     21-MAR-96       45000             E        123 James      12-DEC-78       23000       32000 W        104 Celia      24-OCT-82                   58000 E        105 Robert     15-JAN-84       31000             W        116 Linda      30-JUL-87                   53000 E        117 David      31-DEC-90       78000             W        108 Jode       17-SEP-96       21000       29000 E 7 rows selected. SQL> SQL> SQL> SQL> SQL> SQL> SELECT empno, ename, NVL(curr_salary,44444),   2    RANK()   3    OVER(ORDER BY NVL(curr_salary,44444) desc) salary   4  FROM employee   5  ORDER BY curr_salary desc;      EMPNO ENAME      NVL(CURR_SALARY,44444)     SALARY ---------- ---------- ---------------------- ----------        105 Robert                      44444          3        117 David                       44444          3        122 Alison                      44444          3        104 Celia                       58000          1        116 Linda                       53000          2        123 James                       32000          6        108 Jode                        29000          7 7 rows selected. SQL> SQL> SQL> SQL> SQL> -- clean the table SQL> drop table Employee; Table dropped. SQL> SQL>