Question

我正在寻找T-SQL中的一种方法，以便在数月和数年内获得不同类型成员的数量。我的表：

+-------------+-----------+------------
| Member_type | Join_date |  End_date  |
+-------------+-----------+------------+
| TYPE1       | 12-Nov-07 | 12/11/2015 |
| TYPE2       | 24-Nov-10 | 07/07/2016 |
| TYPE3       | 29-Apr-08 | 28/04/2009 |
| TYPE2       | 28-Apr-06 | 31/03/2007 |
| TYPE1       | 11-Jul-06 | 30/06/2007 |
| TYPE2       | 13-Mar-08 | 12/06/2011 |
+-------------+-----------+------------+

我正在寻找像这样的结果

+-------------+--------+--------+--------+--------+
| Member_type | Jan-15 | Feb-15 | Mar-15 | Apr-15 |
+-------------+--------+--------+--------+--------+
| TYPE1       |      1 |      5 |      4 |      2 |
| TYPE2       |      6 |      4 |      2 |      1 |
| TYPE3       |      5 |      6 |      7 |      8 |
+-------------+--------+--------+--------+--------+

显示近年来在1月，2月等期间给定类型的成员总数。（那些加入的人 - 那些离开的人）

到目前为止，我已经到了这个

SELECT COUNT("membership_type") AS JANUARY2015 
FROM "dbo"."Data" T0 
WHERE (join_date < DATEADD(month, -2, GETDATE())) 
  AND (join_date > GETDATE()) 
GROUP BY T0."membership_type" ;

Answer 1

SELECT Member_type
    , COUNT(CASE WHEN YEAR(join_date) = 2015 AND MONTH(join_date) = 1 THEN 1 END) AS [Jan-15]
    , COUNT(CASE WHEN YEAR(join_date) = 2015 AND MONTH(join_date) = 2 THEN 1 END) AS [Feb-15]
    , COUNT(CASE WHEN YEAR(join_date) = 2015 AND MONTH(join_date) = 3 THEN 1 END) AS [Mar-15]
    , COUNT(CASE WHEN YEAR(join_date) = 2015 AND MONTH(join_date) = 4 THEN 1 END) AS [Apr-15]
    ...
FROM MarketingData
GROUP BY Member_type

T-SQL随着时间的推移，会员类型的多个计数

1 个答案: