Question

我在临时表中有一组值，我想用它来填充三个新列（在下面的MAX(CASE ... END)语句中）。

但是，我发现有时第1列（或第1列和第2列）将包含NULL值，即使第3列有值。

我希望在其他值之后显示的任何NULL值，例如，Col1 = value，Col2 = value，Col3 = NULL。

源代码：

SELECT mc.ID ,mc.Date ,max(case when sv1.row_num = 1 and mc.Date = sv1.Date then sv1.[Col] end) as Col1 ,max(case when sv1.row_num = 2 and mc.Date = sv1.Date then sv1.[Col] end) as Col2 ,max(case when sv1.row_num = 3 and mc.Date = sv1.Date then sv1.[Col] end) as Col3 INTO #Total7 FROM #MyTable3 as mc join ( select #MyTable3.*, row_number() OVER (PARTITION BY [ID] ORDER BY [Date]) as row_num from #MyTable3 )as sv1 on mc.ID = sv1.ID GROUP BY mc.ID ,mc.Date SELECT * FROM #Total7 ORDER BY ID, Date

Answer 1

您的值按此顺序显示，因为NULL is always the first value returned by an ORDER BY，这就是您在ROW_NUMBER()函数中使用的对记录进行排序（确定值是否属于Col1，Col2或Col3）的内容。将您的查询更改为：

SELECT 
     mc.ID 
    ,mc.Date 
    ,max(case when sv1.row_num = 1 and mc.Date = sv1.Date then sv1.[Col] end) as Col1
    ,max(case when sv1.row_num = 2 and mc.Date = sv1.Date then sv1.[Col] end) as Col2
    ,max(case when sv1.row_num = 3 and mc.Date = sv1.Date then sv1.[Col] end) as Col3
INTO #Total7
FROM  
    #MyTable3 as mc  
     join
      (
        select 
            #MyTable3.*,
            ROW_NUMBER() OVER (PARTITION BY [ID] ORDER BY ISNULL([Date], '12/31/9999')) as row_num
        from #MyTable3 
      )as sv1 on 
        mc.ID = sv1.ID 
GROUP BY 
    mc.ID
    ,mc.Date

SELECT * 
FROM #Total7
ORDER BY ID, Date

如果ISNULL()字段为空，[Date]函数将返回远期的日期。因此，任何“真实”日期都会显示第一行数值，任何NULL日期将显示在最后。

或者，如果您不喜欢ISNULL()，则可以使用CASE声明，即

    ROW_NUMBER() OVER 
      (
        PARTITION BY [ID] 
        ORDER BY 
            CASE WHEN [Date] IS NULL THEN 1 ELSE 0 END,
            [Date]
      ) as row_num

Answer 2

为您的加入查询添加日期。

     SELECT 
         mc.ID 
        ,mc.Date 
        ,max(case when sv1.row_num = 1 and mc.Date = sv1.Date then sv1.[Col] end) as Col1
        ,max(case when sv1.row_num = 2 and mc.Date = sv1.Date then sv1.[Col] end) as Col2
        ,max(case when sv1.row_num = 3 and mc.Date = sv1.Date then sv1.[Col] end) as Col3
    INTO #Total7
    FROM  
        #MyTable3 as mc  
         join
          (
            select 
                #MyTable3.*,
                row_number() OVER (PARTITION BY [ID] ORDER BY [Date]) as row_num
            from #MyTable3 
          )as sv1 on 
            mc.ID = sv1.ID 
            and mc.Date = sv1.Date
    GROUP BY 
        mc.ID
        ,mc.Date

    SELECT * 
    FROM #Total7
    ORDER BY ID, Date

在有序集中最后返回NULL值

2 个答案: