Sql server 開窗函數over()的語法

 

用法一：與ROW_NUMBER()函數結合用，給結果進行排序編號，如圖：

 

代碼如下：

SELECT ROW_NUMBER() over(order by RequiredDate) num
       ,* from  [Northwind].[dbo].[Orders]

 

用法二：跟聚合函數一起使用，利用over子句的分組效率比group by子句的效率更高。

在Northwind數據庫的訂單表Orders中查詢"訂單id","客戶id","運費","所有訂單的總數"，“每一個客戶的總運費”，“所有客戶的總運費”，“每一個客戶的平均運費”，“所有客戶的平均運費”,"每一個客戶所有訂單中最大的運費","所有客戶中最大運費","每一個客戶所有訂單中最小的運費","所有客戶中最小運費",如下圖:

 

代碼如下：

  SELECT  [OrderID] --訂單id
      ,[CustomerID] --客戶id
      ,[Freight] --運費
      ,COUNT(OrderID) over() as totalNum --一所有訂單的總數
      ,SUM(Freight) over(partition by customerid) as cusTotalFreight --每一個客戶的總運費
      ,SUM(Freight) over() as totalFreight --所有客戶的總運費
      ,AVG(Freight) over(partition by customerid) as cusAvgFreight --每一個客戶的平均運費
      ,AVG(Freight) over() as avgFreight --所有客戶的平均運費
      ,MAX(Freight) over(partition by customerid) as cusMaxFreight --每一個客戶所有訂單中最大的運費
      ,MAX(Freight) over() as maxFreight --所有客戶中最大運費
      ,MIN(Freight) over(partition by customerid) as cusMinFreight --每一個客戶所有訂單中最小的運費
      ,MIN(Freight) over() as minFreight --所有客戶中最小運費
  FROM [Northwind].[dbo].[Orders]

 

總結：運用開窗函數over()再結合聚合函數，可以使表格向右擴張，並進行一些數據的統計