Question

我正在试图弄清楚如何编写一个计算跨多个列的值的查询，结果表在任何列的每个可能值的每列中都有一个计数。

示例：假设我有 mytable

Source data table:

P1  P2  P3
-----------
a   b   a
a   a   a
b   b   b
a   b   b

我想要一个计算每列中的a和b的查询，产生类似：

Desired query output:

     P1  P2  P3
   -------------
a |  3   1   2
b |  1   3   2

我知道我可以使用组轻松地为单个列执行此操作：

select P1, count(*) as mycounts
from mytable
group by P1

但是每个列都可以这样做吗？

我正在使用SQL Server 2008（T-SQL）。在此先感谢您的帮助！

Answer 1

也许是这样的：

首先是一些测试数据：

DECLARE @tbl TABLE(P1 VARCHAR,P2 VARCHAR,P3 VARCHAR)

INSERT INTO @tbl
SELECT 'a','b','a' UNION ALL
SELECT 'a','a','a' UNION ALL
SELECT 'b','b','b' UNION ALL
SELECT 'a','b','b'

然后像这样的一个支点：

SELECT
    *
FROM
(
    SELECT 'P1' AS P, P1 AS PValue,P1 AS test FROM @tbl
    UNION ALL
    SELECT 'P2',P2,P2 FROM @tbl
    UNION ALL
    SELECT 'P3',P3,P3 FROM @tbl
) AS p
PIVOT
(
    COUNT(PValue)
    FOR P IN ([P1],[P2],[P3])
) AS pvt

Here是有关pivot和unpivot的更多信息

Answer 2

可能效率不高但是有效。

    ;WITH data
AS
(
    SELECT 'a' AS p1, 'b' AS p2, 'a' AS p3
    UNION ALL
    SELECT 'a', 'a','a'
    UNION ALL
    SELECT 'b','b','b'
    UNION ALL
    SELECT 'a','b','b'
)
SELECT
    p_one.value AS header,
    p1,
    p2,
    p3
FROM (SELECT
          p1 AS value,
          count(*) AS p1
      FROM data d
      GROUP BY p1) p_one
left JOIN (SELECT
            p2 AS value,
            count(*) AS p2
            FROM data d
            GROUP BY p2) p_two
    ON p_two.value = p_one.value
left JOIN (SELECT
            p3 AS value,
            count(*) AS p3
            FROM data d
            GROUP BY p3) p_three
    ON p_two.value = p_three.value

SQL Group By / Count：跨多列计算相同的值？

2 个答案: