最近重新又翻看了一下关于sqlserver的书籍，主要查看了一下关于sql中的相关子查询和交叉表方面的知识。
相关子查询和普通子查询区别在于：相关子查询引用了外部查询的列。
这种引用外部查询的能力意味着相关子查询不能自己独立运行，其中对于外部查询引用会使会使其无法正常执行。因此相关子查询的执行顺序如下：
1.首先执行一遍外部查询
2.对于外部查询的每一行分别执行一遍子查询，而且每次执行子查询时候都会引用外部的当前行的值。
使用子查询的结果来确定外部查询的结果集。
举个例子；
SELECT t1.type
FROM titles t1
GROUP BY t1.type
HAVING MAX(t1.advance) >=ALL
(SELECT 2 * AVG(t2.advance)
FROM titles t2
WHERE t1.type = t2.type)
这个结果返回最高预付款超过给定组中平均预付款两倍的书籍类型。
在举个例子：
要求返回每一个编号的最大值(列出id,name,score)
ID Name（编号） Score（分数）
1 a 88
2 b 76
3 c 66
4 c 90
5 b 77
6 a 56
7 b 77
8 c 67
9 a 44
select * from t a where score=
(select Max(Score) from t b where a.name=b.name)
再给一个排位的sql语句
SELECT (
SELECT count(*)+1 as dd
FROM [Test ] as a where a.[F2]<b.[F2] ) AS ord,b.[F1], b.[F2]
FROM [Test ] as b
order by b.[F2];
好了关于sql的相关子查询先讲到这里。
下面说一下交叉表的概念
说到交叉表先提一下递归的select变量
递归的select变量可以使用select语句和子查询将一个变量与其自身拼接起来。
举一个例子
select @var=@var +d.column from table1 a
从而将基础表中垂直的列数据改为水平方向的数据。这样就可以替代游标。
下面就是动态交叉表和静态的交叉表的一个比较，动态的交叉表这样就代替了传统的游标。
交叉表
方法1
select f_nUMBER as '学员',
SUM(case f_subject when 'A01' then f_nUM end) as 'A01',
SUM(case f_subject when 'A02' then f_nUM end) as 'A02' ,
SUM(case f_subject when 'A03' then f_nUM end) as 'A03' ,
SUM(case f_subject when 'A04' then f_nUM end) as 'A04' ,
SUM(case f_subject when 'A05' then f_nUM end) as 'A05' ,
SUM(case f_subject when 'A06' then f_nUM end) as 'A06' ,
SUM(case f_subject when 'A07' then f_nUM end) as 'A07' ,
SUM(case f_subject when 'A08' then f_nUM end) as 'A08' ,
SUM(case f_subject when 'A09' then f_nUM end) as 'A09'
from rowdata group by f_nUMBER order by f_nUMBER
方法2
declare @sql nvarchar(2000)
set @sql=''
select @sql=@sql+ 'sum(case F_subject when '''+ a.F_subject +''' then F_Num else 0 end) as '
+a.F_Name+','
from (select distinct top 100 percent F_subject,F_Name from rowdata b JOIN SUBJECT_name c on b.F_subject=c.F_number order by F_subject ) a
set @sql='select f_nUMBER as '+'"学员",'+@sql + 'count(F_Number) as '+'"考试数目"'+
'from rowdata group by F_Number order by F_Number'
print @sql
exec sp_executesql @sql