Question

这是我原来的表：

customers   shopname    custimport
6247        Bison       2014-05-28  
5753        Brownies    2014-05-28  
14315       Falkners    2014-05-28  
6392        Bison       2014-08-31  
5483        Brownies    2014-08-31  
14395       Falkners    2014-08-31

在我的报告中，我制作了这个矩阵

           [custimport]
[shopname] [Sum(customers)]

我知道只有2个custimport日期，我需要总和（客户）之间的差异。基本上，我需要找到商店在2个进口日期之间获得/丢失的客户数量。我不确定用什么表达来实现这个目标。

作为参考，这是现在的矩阵：

       5/28/2014    8/31/2014
Bison       6247    6392
Brownies    5753    5483
Falkners    14315   14395

这就是我想要的结果：

       5/28/2014    8/31/2014   diff
Bison       6247    6392        145
Brownies    5753    5483        -270
Falkners    14315   14395       80

Answer 1

以下内容应该是矩阵的数据集。

    SELECT ORI.SHOPNAME,
           ORI.CUSTIMPORT,
           B.CUSTOMERS-A.CUSTOMERS AS DIFF
    FROM
      (SELECT SHOPNAME,
              CUSTIMPORT
       FROM CUSTOMER_TABLE) ORI
    INNER JOIN
      (SELECT CUSTOMERS,
              SHOPNAME,
              CUSTIMPORT
       FROM CUSTOMER_TABLE)A ON ORI.SHOPNAME=A.SHOPNAME
    AND A.CUSTIMPORT=ORI.CUSTIMPORT
    INNER JOIN
      (SELECT CUSTOMERS,
              SHOPNAME,
              CUSTIMPORT
       FROM CUSTOMER_TABLE)B ON A.SHOPNAME=B.SHOPNAME
    AND A.CUSTIMPORT <> B.CUSTIMPORT
    AND ORI.SHOPNAME=B.SHOPNAME
    ORDER BY A.CUSTIMPORT

Answer 2

编辑：

这个表达对我有用：

=Sum(Fields!Customers.Value) - Previous(Sum(Fields!Customers.Value))

仅限SQL：

SELECT [shopname], [2014-05-28], [2014-08-31], [2014-05-28] - [2014-08-31] 'Diff'
FROM #Test
PIVOT
(
Sum(customers)
FOR custimport in ([2014-05-28], [2014-08-31])
) AS PVT_Table

矩阵中列之间的差异

2 个答案: