pivot的用法，一看就会

最近发现之前在学校学习的sql语句好像基本没啥技术含量
最近发现SqlServer中提供的一个关键字居然从来没用过

 
select (p.LastName + ', ' + p.FirstName) FullName,
MONTH(sh.OrderDate) Month,
count(SalesOrderID) TotalOrder
from Sales.SalesOrderHeader sh
join Person.Person p
on sh.SalesPersonID = p.BusinessEntityID
where month(OrderDate) between 1 and 5
group by p.LastName + ', ' + p.FirstName, MONTH(sh.OrderDate)
order by FullName;

先不管他这组数据具体要干啥
首先需要明白 group by 和 PIVOT 都具有分组功能，两者似乎只能选择其中之一

这里是按照名字和月份分组
下面使用povot进行分组

 select FullName ,[1] as Jan,[2] as Feb,[3] as Mar,[4] as Apr,[5] as May from (
 select  (p.LastName + ', ' + p.FirstName) FullName,
 MONTH(sh.OrderDate) Month ,
SalesOrderID
from Sales.SalesOrderHeader sh
join Person.Person p
on sh.SalesPersonID = p.BusinessEntityID
where month(OrderDate) between 1 and 5 ) c 
PIVOT(count(SalesOrderID) FOR Month  IN([1],[2],[3],[4],[5]))
AS T 

这里显然是将
count(SalesOrderID) FOR Month IN([1],[2],[3],[4],[5])
将计算好的count数据进行按for后面的月份分组
比如
一月份多少多少
二月份多少多少
听其他博主说了一堆，把注意力全都集中在转列上，这个明明是分组，只是换了一种方式，不然要你在前面写个聚合函数干啥，这里就可以看出是根据销售数量按月分组
再比如下面这道题

select CustomerID, year(OrderDate) Year, sum(sd.OrderQty) TotalPurchaseQuantity
from Sales.SalesOrderHeader sh
join sales.SalesOrderDetail sd
on sh.SalesOrderID = sd.SalesOrderID
where CustomerID between 18010 and 18014
group by CustomerID, year(OrderDate)
order by year;

需要我们将这个格式的输出装换为下面这个格式

select year, isnull([18010],0) as[18010] ,isnull([18011],0) as [18011],isnull([18014],0) as [18014],isnull([18012],0) as [18012],isnull([18013],0)as [18013] from (
select CustomerID,
year(OrderDate) Year,
OrderQty
from Sales.SalesOrderHeader sh
join sales.SalesOrderDetail sd
on sh.SalesOrderID = sd.SalesOrderID
where CustomerID between 18010 and 18014
) as c
pivot (sum(OrderQty) for CustomerID in  ([18010],[18011],[18014],[18012],[18013]) ) s

所以在此证明这就是个分组排序而已