0

我在下面有 3 个选择语句，每个语句创建完全相同的计算列，然后我计算该计算列中的所有是和所有否。目前它显示该区域 3 次，旁边有 3 个结果。我希望它显示该区域一次，其中包含 3 个带有总计的新列，所以目前

我有： 

region1 computedCol1

region1 computedCol2

region1 computedCol3

region2 computedCol1

region2 computedCol2

region2 computedCol3

我想：

Region1,  computedCol1, computedCol2, computedCol3

Region2,  computedCol1, computedCol2, computedCol3

Region3,  computedCol1, computedCol2, computedCol3

如果我使用“union all”，我会得到：

region1 computedCol1

region2 computedCol1

region3 computedCol1



SELECT a.region, COUNT(*) AS [computedCol1]    
(
SELECT  DISTINCT table1.serial1, table1.serial2, region,
    CASE WHEN table2.serial1 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol1],
    CASE WHEN table3.serial2 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol2],
    CASE WHEN table3.serial2 IS NULL AND table2.serial1 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol3]

FROM    table1
        LEFT JOIN table2
            ON table2.serial1 = table1.serial1
        LEFT JOIN dbo.EPG
            table3.serial2 = table1.serial2
)a
    WHERE region in (
        '37000','38000','39000','41000','42000','43000','44000','46000','45000','51000','52000','53000','54000',
        '55000','56000','57000','58000','59000','61000','62000','63000','64000','65000','66000','67000','68000',
        '69000','30000','33000','36000','34000','35000','31000','32000','N/A'   )
and [CCA Match Org] in  ('no')
    GROUP BY a.region

union

SELECT b.region, COUNT(*) AS [computedCol2], region   
(
SELECT  DISTINCT table1.serial1, table1.serial2,
    CASE WHEN table2.serial1 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol1],
    CASE WHEN table3.serial2 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol2],
    CASE WHEN table3.serial2 IS NULL AND table2.serial1 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol3]

FROM    table1
        LEFT JOIN table2
            ON table2.serial1 = table1.serial1
        LEFT JOIN dbo.EPG
            table3.serial2 = table1.serial2
)b
    WHERE region in     (
        '37000','38000','39000','41000','42000','43000','44000','46000','45000','51000','52000','53000','54000'
        '55000','56000','57000','58000','59000','61000','62000','63000','64000','65000','66000','67000','68000',
        '69000','30000','33000','36000','34000','35000','31000','32000','N/A'   )
and [CCA Match Org] in ('yes')
    group by b.region

union

SELECT c.region, COUNT(*) AS [computedCol3], region   
(
SELECT  DISTINCT table1.serial1, table1.serial2,
    CASE WHEN table2.serial1 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol1],
    CASE WHEN table3.serial2 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol2],
    CASE WHEN table3.serial2 IS NULL AND table2.serial1 IS NULL THEN 'No' ELSE 'Yes' END AS [computedCol3]

FROM    table1
        LEFT JOIN table2
            ON table2.serial1 = table1.serial1
        LEFT JOIN dbo.EPG
            table3.serial2 = table1.serial2
)c
    WHERE region in     (
        '37000','38000','39000','41000','42000','43000','44000','46000','45000','51000','52000','53000','54000'
        '55000','56000','57000','58000','59000','61000','62000','63000','64000','65000','66000','67000','68000',
        '69000','30000','33000','36000','34000','35000','31000','32000','N/A'   )
and [CCA Match Org] in ('yes', 'no')
    group by c.region
4

2 回答 2

0

这就是 UNION 在所有数据库中的工作方式。它不会创建新列，而是包含所有行。

将行转换为列是使用PIVOT运算符完成的。本质上，PIVOT 按非透视列（在本例中为区域）对行进行分组，并将聚合应用于每个生成的列。

于 2012-06-01T14:38:24.787 回答
0

您可以像其他答案建议的那样使用 pivot ，也可以将 SUM 与 CASE 结合使用来获得相同的结果。如果你有很多computedcols，这会变得很痛苦。

我在我的例子中使用了 Oracle，但这个概念应该转化为 SQLServer。

例子：

SELECT 'A' region, 1 col_num, 10 computed_value FROM DUAL
UNION
SELECT 'A' region, 2 col_num, 11 computed_value FROM DUAL
UNION
SELECT 'A' region, 3 col_num, 12 computed_value FROM DUAL
UNION
SELECT 'B' region, 1 col_num, 20 computed_value FROM DUAL
UNION
SELECT 'B' region, 2 col_num, 21 computed_value FROM DUAL
UNION
SELECT 'B' region, 3 col_num, 22 computed_value FROM DUAL

这将返回：

REGION  COL_NUM COMPUTED_VALUE
A           1   10
A           2   11
A           3   12
B           1   20
B           2   21
B           3   22

如果您按如下方式嵌套该查询：

SELECT     region,
           SUM (CASE col_num WHEN 1 THEN computed_value ELSE NULL END) compulated_1,
           SUM (CASE col_num WHEN 2 THEN computed_value ELSE NULL END) compulated_2,
           SUM (CASE col_num WHEN 3 THEN computed_value ELSE NULL END) compulated_3
    FROM   (SELECT 'A' region, 1 col_num, 10 computed_value FROM DUAL
            UNION
            SELECT 'A' region, 2 col_num, 11 computed_value FROM DUAL
            UNION
            SELECT 'A' region, 3 col_num, 12 computed_value FROM DUAL
            UNION
            SELECT 'B' region, 1 col_num, 20 computed_value FROM DUAL
            UNION
            SELECT 'B' region, 2 col_num, 21 computed_value FROM DUAL
            UNION
            SELECT 'B' region, 3 col_num, 22 computed_value FROM DUAL)
GROUP BY   region

给你：

REGION  COMPULATED_1    COMPULATED_2    COMPULATED_3
A           10          11           12
B           20          21           22
于 2012-06-01T17:40:53.343 回答