替代爲「ROW_NUMBER（）以上（分區由」

Question

尋找低於
替代了查詢，而不row_number() over(partition by：

select *,row_number() over(partition by [Tag Name],[PRD] order by [Start Time]) num 
from [PRD Project].[Data].[efan1] 

Answer 1

我們可以嘗試用標識列創建tablevariable並插入選定的數據得到ROW_NUMBER作爲輸出

declare @test table (i int identity(1,1), test varchar(10)) 

Answer 2

雖然我覺得ROW_NUMBER可能是做到這一點的最佳方式，這裏有一個替代方案。

的設置：

CREATE TABLE [#test] ([Col1] CHAR(1) -- This is what we're partitioning by 
        , [Col2] INT); -- This is what we're ordering by 

INSERT INTO [#test] 
SELECT CHAR(ABS(CHECKSUM(NEWID()))%16+65) -- Random single letters A:P 
     , ABS(CHECKSUM(NEWID())) % 1000  -- Random numbers 1:1000 
FROM [sys].[objects]; 

的查詢：

-- Original ROW_NUMBER query 
SELECT * 
     , ROW_NUMBER() OVER(PARTITION BY [Col1] ORDER BY [Col2]) AS [RN] 
FROM [#test] 
ORDER BY [Col1] 
     , [rn]; 


-- Alternative w/ CROSS APPLY 
SELECT * 
FROM [#test] [t1] 
     -- For each row in #test, 
     -- Get the number of rows in #test where our partitioning column (Col1) is the same and the ordering column (Col2) is lower 
     -- Add 1 to get the initial value to start at 1 instead of 0 (i.e. for the first value per partition, there are 0 rows with a lower ordering value) 
     CROSS APPLY (SELECT COUNT(*) + 1 AS [rn] 
        FROM [#test] [t2] 
        WHERE [t1].[Col1] = [t2].[Col1] 
         AND [t1].[Col2] > [t2].[Col2]) [t2] 
ORDER BY [Col1] 
     , [rn];