Question

我正在查询一张基本上是一堆日期的表格。

FRECDCUST              DTRECDCUST              DTINSPECTED             DTRECDCUSTPO           DTADDEDSO                 DTSHIPPEDSUP                  DTQUOTEDCUST
1/1/1900 12:00:00 AM    7/27/2010 12:00:00 AM   7/30/2010 12:00:00 AM   7/26/2010 12:00:00 AM   8/6/2010 12:00:00 AM    1/1/1900 12:00:00 AM    8/6/2010 12:00:00 AM
1/1/1900 12:00:00 AM    7/27/2010 12:00:00 AM   7/30/2010 12:00:00 AM   7/26/2010 12:00:00 AM   8/6/2010 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM
1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM
1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM
1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM    1/1/1900 12:00:00 AM
1/1/1900 12:00:00 AM    3/12/2010 12:00:00 AM   3/15/2010 12:00:00 AM   3/11/2010 12:00:00 AM   3/16/2010 12:00:00 AM   3/24/2010 12:00:00 AM   4/13/2010 12:00:00 AM
1/1/1900 12:00:00 AM    11/6/2009 12:00:00 AM   11/6/2009 12:00:00 AM   11/3/2009 12:00:00 AM   11/9/2009 12:00:00 AM   11/20/2009 12:00:00 AM  12/7/2009 12:00:00 AM

我正在考虑每行上两个日期之间的日期差异。例如，检查的日期如下：

SELECT (CASE
          WHEN year(SYC_ext.DTRECDCUST) <> 1900
               AND year(SYC_ext.DTINSPECTED) <> 1900
          THEN
            DATEDIFF(dd, SYC_ext.DTRECDCUST, SYC_ext.DTINSPECTED)
          ELSE
            NULL
        END)
         AS [AverageDaystoInspect]
Blah blah...

这会创建一行不同的DateDiff计算，最终看起来像这样：

Product Class   Average Days to Inspect Average Days to Process SO  Average Days to Ship to Affiliate   Average Days to Quote   Average Days to Receive Ammended PO
01  2   4   8   27  21
01  2   4   8   27  21
01  3   5   2   88  8
01  3   5   2   88  8
01  1   8   11  72  1
01  3   2   7   27  0
01  5   2   7   27  0

请原谅间距问题。

因此，数据集中的每一行都以一个数字开头，该数字表示此情况下的产品类别01，然后是一系列日计算。

我可以平均这些值，并为每个产品类别的每个类别提供平均值。但是，他们希望在表格或矩阵的左侧进行日计算，产品类别位于顶部，与此处的完全相反。如果我在Excel中转动它们，我会得到以下结果：

                          01    05     10
Avg Days to Inspect       5      3      5
Avg Days to Process PO    4      5      5
Avg Days to Ship          20     18     19

我希望这是有道理的。知道如何在SSRS或t-sql中执行此操作吗？

Answer 1

这样做，假设您可以将第一个查询中的数据加载到中间临时表中：

DECLARE @t TABLE
    (
      ProductClass INT
    , InspectDays INT
    , ProcessDays INT
    , ShipDays INT
    )

INSERT  INTO @t
        SELECT  1
              , 2
              , 4
              , 8
        UNION ALL
        SELECT  1
              , 4
              , 5
              , 2
        UNION ALL
        SELECT  5
              , 2
              , 4
              , 8
        UNION ALL
        SELECT  10
              , 3
              , 5
              , 2

SELECT  'Avg Inspect'
      , [01] = AVG(CASE WHEN ProductClass = 1 THEN InspectDays
                   END)
      , [05] = AVG(CASE WHEN ProductClass = 5 THEN InspectDays
                   END)
      , [10] = AVG(CASE WHEN ProductClass = 10 THEN InspectDays
                   END)
FROM    @t
UNION ALL
SELECT  'Avg Process'
      , [01] = AVG(CASE WHEN ProductClass = 1 THEN ProcessDays
                   END)
      , [05] = AVG(CASE WHEN ProductClass = 5 THEN ProcessDays
                   END)
      , [10] = AVG(CASE WHEN ProductClass = 10 THEN ProcessDays
                   END)
FROM    @t

如何使用SQL2000转动/取消某些东西

1 个答案: