Sql Server classic: the ranks of the conversion and perspective (Pivot Method)

Sql server common packet aggregation very broad application, but sometimes we need to use a similar excel show the effect of perspective;

So how do you achieve this perspective effect?

In the following example:

To Oracle database scott classic example:

　　We require each department to show the next (deptno), functional (job) of all comprehensive salary is how much, under normal circumstances we will use the following query

1 select deptno,job,SUM(sal) as sum_sal
2     from emp
3     group by deptno,job

Results are as follows:

 

But if we ask deptno vertical arrangement, Job horizontal arrangement should be?

Conventional worded as follows:

1 --常规透视
2 select deptno,
3     sum(case  when job='ANALYST' then sal else 0 end) as ANALYST,
4     sum(case  when job='CLERK' then sal else 0 end) as CLERK,
5     sum(case  when job='MANAGER' then sal else 0  end) as MANAGER,
6     sum(case  when job='PRESIDENT' then sal  else 0 end) as PRESIDENT,
7     sum(case  when job='SALESMAN' then sal else 0  end) as SALESMAN
8     from emp 
9     group by deptno

 

The results obtained were as follows:

 

Used to form the operation of such arrangement might feel a little more intuitive;

当然这是常规的写法，Sql Server 自带的Pivot方法也可以实现：不过对版本的要求至少是2008

如下：

1 --pivot透视    
2     select deptno,ANALYST,CLERK,MANAGER,PRESIDENT,SALESMAN
3         from
4             (select deptno,job,sal from emp)as a
5         pivot (sum(sal) 
6             for job 
7                 in (ANALYST,CLERK,MANAGER,PRESIDENT,SALESMAN)) as b

效果同上

 

 

但是需要注意一点，如果这个查询的行和列相反(即：deptno横向，job纵向)，

因为deptno值全是数值，那么在书写的时候要注意一下，加上"[]"，

如下

1 select job,[10],[20],[30]
2     from 
3         (select job,deptno,sal from emp) a
4     pivot (sum(sal) 
5         for deptno 
6             in ([10],[20],[30]) )b

效果如下：

 

 

这个是透视，逆透视--未完待续…………