GROUP BY的扩展

GROUP BY的扩展主要包括ROLLUP，CUBE，GROUPING SETS三种形式。

ROLLUP

rollup相对于简单的分组合计增加了小计和合计，解释起来会比较抽象，下面我们来看看具体事例。

例1，统计不同部门工资的总和和所有部门工资的总和。

SQL> select deptno,sum(sal) from emp group by rollup(deptno);

    DEPTNO   SUM(SAL)
---------- ----------
        10       8750
        20      10875
        30       9400
                29025

例2，该例中先对deptno进行分组，再对job进行分组

SQL> select deptno,job,sum(sal) from emp group by rollup(deptno,job);

    DEPTNO JOB           SUM(SAL)
---------- --------- ----------
        10 CLERK           1300    --10号部门中JOB为CLERK的工资的总和
        10 MANAGER         2450
        10 PRESIDENT       5000
        10                 8750    --10号所有工种工资的总和
        20 CLERK           1900
        20 ANALYST         6000
        20 MANAGER         2975
        20                10875
        30 CLERK            950
        30 MANAGER         2850
        30 SALESMAN        5600
        30                 9400
                          29025   --所有部门，所有工种工资的总和
13 rows selected.

如果要用普通的分组函数实现，可用UNION ALL语句：

--实现单个部门，单个工种的工资的总和
select deptno,job,sum(sal) from emp group by deptno,job
union all
--实现单个部门工资的总和
select deptno,null,sum(sal) from emp group by deptno
union all
--实现所有部门工资的总和
select null,null,sum(sal) from emp
order by 1,2

下面我们分别来看看两者的执行计划及统计信息，

ROLLUP语句：

Execution Plan
-----------------------------------------------------------------------------
| Id  | Operation         | Name | Rows  | Bytes | Cost (%CPU)| Time     |
-----------------------------------------------------------------------------
|   0 | SELECT STATEMENT     |        |     11 |    132 |      3  (34)| 00:00:01 |
|   1 |  SORT GROUP BY ROLLUP|        |     11 |    132 |      3  (34)| 00:00:01 |
|   2 |   TABLE ACCESS FULL  |   EMP  |     14 |    168 |      2   (0)| 00:00:01 |
-----------------------------------------------------------------------------
Statistics
----------------------------------------------------------
      0  recursive calls
      0  db block gets
      2  consistent gets
      0  physical reads
      0  redo size
    895  bytes sent via SQL*Net to client
    519  bytes received via SQL*Net from client
      2  SQL*Net roundtrips to/from client
      1  sorts (memory)
      0  sorts (disk)
     13  rows processed

UNION ALL语句：

Execution Plan
-----------------------------------------------------------------------------
| Id  | Operation         | Name | Rows  | Bytes | Cost (%CPU)| Time     |
-----------------------------------------------------------------------------
|   0 | SELECT STATEMENT     |        |     15 |    150 |      9  (34)| 00:00:01 |
|   1 |  SORT ORDER BY       |        |     15 |    150 |      8  (75)| 00:00:01 |
|   2 |   UNION-ALL          |        |        |        |             |          |
|   3 |    HASH GROUP BY     |        |     11 |    132 |      3  (34)| 00:00:01 |
|   4 |     TABLE ACCESS FULL|   EMP  |     14 |    168 |      2   (0)| 00:00:01 |
|   5 |    HASH GROUP BY     |        |      3 |     15 |      3  (34)| 00:00:01 |
|   6 |     TABLE ACCESS FULL|   EMP  |     14 |     70 |      2   (0)| 00:00:01 |
|   7 |    SORT AGGREGATE    |        |      1 |      3 |             |          |
|   8 |     TABLE ACCESS FULL|   EMP  |     14 |     42 |      2   (0)| 00:00:01 |
-----------------------------------------------------------------------------

Statistics
----------------------------------------------------------
      0  recursive calls
      0  db block gets
      6  consistent gets
      0  physical reads
      0  redo size
    895  bytes sent via SQL*Net to client
    519  bytes received via SQL*Net from client
      2  SQL*Net roundtrips to/from client
      1  sorts (memory)
      0  sorts (disk)
     13  rows processed

不难看出，相同的功能实现，ROLLUP相对于UNION ALL效率有了极大的提升。

CUBE

cube相对于rollup，结果输出更加详细。

例1，在本例中还不是很明显。

SQL> select deptno,sum(sal) from emp group by cube(deptno);

    DEPTNO   SUM(SAL)
---------- ----------
                29025
        10       8750
        20      10875
        30       9400

例2，相对于rollup，cube还对工种这一列进行了专门的汇总。

SQL> select deptno,job,sum(sal) from emp group by cube(deptno,job);

    DEPTNO JOB           SUM(SAL)
---------- --------- ----------
                          29025
           CLERK           4150
           ANALYST         6000
           MANAGER         8275
           SALESMAN        5600
           PRESIDENT       5000
       10                  8750
       10  CLERK           1300
       10  MANAGER         2450
       10  PRESIDENT       5000
       20                 10875
       20  CLERK           1900
       20  ANALYST         6000
       20  MANAGER         2975
       30                  9400
       30  CLERK            950
       30  MANAGER         2850
       30  SALESMAN        5600
18 rows selected.

GROUPING SETS

GROUPING SETS相对于ROLLUP和CUBE，结果是分类统计的，可读性更好一些。

例1：

SQL> select deptno,job,to_char(hiredate,'yyyy')hireyear,sum(sal) from emp group by grouping sets(deptno,job,to_char(hiredate,'yyyy'));

    DEPTNO JOB         HIRE   SUM(SAL)
---------- ---------   ---- ----------
           CLERK                  4150
           SALESMAN               5600
           PRESIDENT              5000
           MANAGER                8275
           ANALYST                6000
       30                         9400
       20                        10875
       10                         8750
                        1987      4100
                        1980       800
                        1982      1300
                        1981     22825

例2：

SQL> select deptno,job,sum(sal) from emp group by grouping sets(deptno,job);

    DEPTNO JOB           SUM(SAL)
---------- ---------   ----------
           CLERK             4150
           SALESMAN          5600
           PRESIDENT         5000
           MANAGER           8275
           ANALYST           6000
        30                   9400
        20                  10875
        10                   8750
8 rows selected.

对于该例，如何用UNION ALL实现呢？

select null deptno,job,sum(sal) from emp group by job
union all
select deptno,null,sum(sal) from emp group by deptno;

两者的执行计划及统计信息分别如下：

GROUPING SETS：

Execution Plan
--------------------------------------------------------------------------------------------------------
| Id  | Operation                   | Name                      | Rows  | Bytes | Cost (%CPU)| Time     |
--------------------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT            |                           |    11 |   352 |    10  (20)| 00:00:01 |
|   1 |  TEMP TABLE TRANSFORMATION  |                           |       |       |            |          |
|   2 |   LOAD AS SELECT            | SYS_TEMP_0FD9D6795_E71F79 |       |       |            |          |
|   3 |    TABLE ACCESS FULL        | EMP                       |    14 |   168 |     2   (0)| 00:00:01 |
|   4 |   LOAD AS SELECT            | SYS_TEMP_0FD9D6796_E71F79 |       |       |            |          |
|   5 |    HASH GROUP BY            |                           |     1 |    19 |     3  (34)| 00:00:01 |
|   6 |     TABLE ACCESS FULL       | SYS_TEMP_0FD9D6795_E71F79 |     1 |    19 |     2   (0)| 00:00:01 |
|   7 |   LOAD AS SELECT            | SYS_TEMP_0FD9D6796_E71F79 |       |       |            |          |
|   8 |    HASH GROUP BY            |                           |     1 |    26 |     3  (34)| 00:00:01 |
|   9 |     TABLE ACCESS FULL       | SYS_TEMP_0FD9D6795_E71F79 |     1 |    26 |     2   (0)| 00:00:01 |
|  10 |   VIEW                      |                           |     1 |    32 |     2   (0)| 00:00:01 |
|  11 |    TABLE ACCESS FULL        | SYS_TEMP_0FD9D6796_E71F79 |     1 |    32 |     2   (0)| 00:00:01 |
--------------------------------------------------------------------------------------------------------
Statistics
----------------------------------------------------------
      4  recursive calls
     24  db block gets
     17  consistent gets
      3  physical reads
   1596  redo size
    819  bytes sent via SQL*Net to client
    519  bytes received via SQL*Net from client
      2  SQL*Net roundtrips to/from client
      0  sorts (memory)
      0  sorts (disk)
      8  rows processed

UNION ALL：

----------------------------------------------------------------------------
| Id  | Operation           | Name  | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------
|   0 | SELECT STATEMENT    |       |     8 |    65 |     6  (67)| 00:00:01 |
|   1 |  UNION-ALL          |       |       |       |            |          |
|   2 |   HASH GROUP BY     |       |     5 |    50 |     3  (34)| 00:00:01 |
|   3 |    TABLE ACCESS FULL|  EMP  |    14 |   140 |     2   (0)| 00:00:01 |
|   4 |   HASH GROUP BY     |       |     3 |    15 |     3  (34)| 00:00:01 |
|   5 |    TABLE ACCESS FULL|  EMP  |    14 |    70 |     2   (0)| 00:00:01 |
----------------------------------------------------------------------------

Statistics
----------------------------------------------------------
      0  recursive calls
      0  db block gets
      4  consistent gets
      0  physical reads
      0  redo size
    819  bytes sent via SQL*Net to client
    519  bytes received via SQL*Net from client
      2  SQL*Net roundtrips to/from client
      0  sorts (memory)
      0  sorts (disk)
      8  rows processed

和rollup不同的是，grouping sets的效率竟然比同等功能的union all语句低，这实现有点出乎意料。看来，也不可盲目应用Oracle提供的方案，至少，在本例中是如此。