这个查询出现次数的SQL语句怎么写才好？

[onlyonehaimeika]

对数据库
tid items
T100 I1,I2,I5
T200 I2,I4
T300 I2,I3
T400 I1,I2,I4
T500 I1,I3
T600 I2,I3
T700 I1,I3
T800 I1,I2,I3,I5
T900 I1,I2,I3
想要查询各项组合出现的次数，即以下结果的话，SQL语句应该怎么写才好。请各位发表高见
items count
I1,I2 4
I1,I3 3
I1,I4 1
I1,I5 2
I2,I3 4
I2,I5 2
I3,I5 1

[onlyonehaimeika]

谢谢‘网中漫步’,在你的帮助下二维查询已经成功的运行了。

[webswim]

建立一张临时表tmp1(fld1)，存储I1~I5, 用查询提取二维组合：
( select a.fld1 I_first, b.fld1 I_second
from tmp1 a, (select * from tmp1) b
where a.fld1 <> b.fld1
) x

然后判断上述二维组合在你的表（sourceData）中出现的次数：
select x.I_first + ',' + x.I_second as items, count(*) as count
from
( select a.fld1 I_first, b.fld1 I_second
from tmp1 a, (select * from tmp1) b
where a.fld1 <> b.fld1
) x, sourceData y
where b.items like '%' + x.I_first + '%'
and b.items like '%' + x.I_second + '%'
group by x.I_first, x.I_second

希望能有所帮助

[wawaren]

这个就是查询所有长度为二的关联规则，可以通过该表的一个字连接来实现。

select a.itmes,b.items,count(*)
from a,b

[onlyonehaimeika]

Thank you,webswim.I am trying to run it.
And thank you wawaren too.Yes ,It is a part of association rules mining.But I did not understand what you said very well.What does 'a,b'mean? Or what is in 'a,b'?Would you please say that more clearly?
I am very sorry to write in English.Because now,I can only read in Chanese but can not write.Really hope your helpness! Thank you.

[onlysgirl]

select count(1) from table group by items

[littlecpu]

不用在一棵树上吊死，求速用存储，不求速拉到开发工具中用高级语言统计

[ckp]

gz