SQL通过滑动窗口保留发生次数

Question

在上一个问题中，我需要计算一个变量出现的次数。

提供的代码如下：

SELECT 
    [Date], Code, 
    [Count] = COUNT(*) OVER (PARTITION BY Code ORDER BY [Date] ROWS UNBOUNDED PRECEDING)
FROM dbo.YourTable
ORDER BY [Date];

但是，现在我需要对该代码进行改进：

假设我有下表：

   Date          | Code
   ------------------------
   2010/01/01    | 25
   2010/01/01    | 22
   2010/01/01    | 23
   2010/01/01    | 25
   2010/01/02    | 23
   2010/01/02    | 23
   2010/01/03    | 23
   2010/01/04    | 23
   2010/01/05    | 23
   2010/01/06    | 23
   2010/01/07    | 23
   .....
   2013/03/02    | 21

现在，我需要计算特定时间段内的出现次数。期望的输出如下（为简单起见，假设时间范围为n = 2天）

    Date         | Code  |  Occurrences
   ------------------------------------
   2010/01/01    | 25    |      1
   2010/01/01    | 22    |      1
   2010/01/01    | 23    |      1
   2010/01/01    | 25    |      2
   2010/01/02    | 23    |      2
   2010/01/02    | 23    |      3
   2010/01/03    | 23    |      3 -> We are not considering the occurence in 2011/01/01 as it is out of the scope now
   2010/01/04    | 23    |      2 -> Considers only occurrences in 01/03 and 01/04 
   2010/01/05    | 23    |      2
   2010/01/06    | 23    |      2
   2010/01/07    | 23    |      2
   .....
   2013/03/02    |  21   |      1

也就是说，我需要知道在过去的“ n”个月中，代码“ x”在我的表中出现了多少次。

它在SQL Server 2012中运行。

先感谢您。

Answer 1

将选项与CTE，CROSS
APPLY
运算符和ROW_NUMBER排序功能一起使用

 ;WITH cte AS
 (
  SELECT [Date], Code
  FROM dbo.YourTable  
  GROUP BY [Date], Code
  )
  SELECT c.Date, c.Code, o.Occurrences
  FROM cte c 
    CROSS APPLY (
                 SELECT t2.[Date], t2.Code,
                        ROW_NUMBER() OVER(PARTITION BY c.[Date] ORDER BY t2.[Date]) AS Occurrences        
                 FROM dbo.YourTable t2
                 WHERE c.Code = t2.Code
                   AND DATEDIFF(day, t2.[Date], c.[Date]) BETWEEN 0 AND 1
                 ) o
  WHERE c.Code = o.Code AND c.[Date] = o.[Date] 
  ORDER BY c.[Date]

关于 SQLFiddle的 演示

为了提高性能，请使用此索引

CREATE INDEX x ON dbo.YourTable([Date], Code)