如何在T-SQL语句中找到Apriori关联？

Question

数据库：

Transaction  ProductID
 1              1000
 2              1000
 2              1001 
 3              1000
 3              1002
 4              1000
 4              1001
 5              1003

上表中，如何用T-SQL语句找到这个结果？

+-------------+-----------+-----------------+
| PRODUCTID1  | PRODUCTID2|     SUPPORT     |
+-------------+-----------+-----------------+
|      1000   |   1001    |         2       |
+-------------+-----------+-----------------+
|      1000   |   1002    |         1       |
+-------------+-----------+-----------------+
|      1000   |   1003    |         0       |
+-------------+-----------+-----------------+
|      1001   |   1002    |         0       |
+-------------+-----------+-----------------+
|      1001   |   1003    |         0       |
+-------------+-----------+-----------------+
|      1002   |   1003    |         0       |,
+-------------+-----------+-----------------+

测试表：

create table transactions(
   ORDERID    INT,
   PRODUCTID INT
);

insert into transactions(ORDERID, PRODUCTID)
values ('1', '1000')
      ,('2', '1000')
      ,('2', '1001')
      ,('3', '1000')
      ,('3', '1002')
      ,('4', '1000')
      ,('4', '1001'),
       ('5', '1003');

提前致谢。

Answer 1

你能推断出零吗？即我们可以说没有列出的任何内容都是零吗？如果是这样：

select t1.PRODUCTID as [PRODUCTID1], t2.PRODUCTID as [PRODUCTID2],
       COUNT(1) as [SUPPORT]
from transactions t1 inner join transactions t2
  on t1.ORDERID = t2.ORDERID
  and t1.PRODUCTID < t2.PRODUCTID
group by t1.PRODUCTID, t2.PRODUCTID

输出：

PRODUCTID1  PRODUCTID2  SUPPORT
----------- ----------- -----------
1000        1001        2
1000        1002        1