如何找到没有唯一ID的两个查询之间的差异

我正在使用以下查询来尝试查找两个数据库的查询结果之间的差异。我无法将AccountID用作比较点，因为它们不在不同数据库之间排列。 ctes中的选择都返回200105行，但cte上的连接结果具有201779行。这种情况下的数据应该能够匹配。我在这条正确的轨道上？如果是的话，我错过了什么，或者我应该寻找一种不同的方法。如何找到没有唯一ID的两个查询之间的差异

with cteDev3 (AccountID, SecurityID, UniqueAccountID, BuyPrice, BuyDate, Shares)as 
     (select Al.AccountID, al.SecurityID,ad.UniqueAccountId, BuyPrice, BuyDate, Shares 
     from dev3TMAd.dbo.AccountLot al 
     join dev3TMAd.dbo.AccountDetails ad 
     on al.AccountID = ad.AccountId 
     where ad.EnterpriseId = 'HuberFinancial'), 
    cteTest2 (AccountID, SecurityID, UniqueAccountID, BuyPrice, BuyDate, Shares)as 
     (select Al.AccountID, al.SecurityID,ad.UniqueAccountId, BuyPrice, BuyDate, Shares 
     from test2TMAd.dbo.AccountLot al 
     join test2TMAd.dbo.AccountDetails ad 
     on al.AccountID = ad.AccountId 
     where ad.EnterpriseId = 'HuberFinancial') 

select c3.UniqueAccountId as 'Dev3' , t2.UniqueAccountId as 'Test2' 
from cteTest2 t2 
join cteDev3 c3 
    on c3.UniqueAccountId = t2.UniqueAccountId 
    and c3.SecurityID = t2.SecurityID 
    and c3.BuyDate = t2.BuyDate 
    and c3.Shares = t2.Shares 
    and c3.BuyPrice = t2.BuyPrice

这是结果集的样子对于那些在热膨胀系数的部分子查询：

AccountID | SecurityID | UniqueAccountId | BuyPrice | BuyDate | Shares 
949662 | 17030 | 11284035 | 42.690021 | 2007-12-19 00:00:00 | 4.710000 
949662 | 17030 | 11284035 | 42.690629 | 2007-12-19 00:00:00 | 13.521000 
949662 | 17030 | 11284035 | 42.611940 | 2007-12-19 00:00:00 | 0.134000 
949662 | 17030 | 11284035 | 39.323467 | 2008-03-10 00:00:00 | 0.946000 
949662 | 17030 | 11284035 | 40.831884 | 2008-06-10 00:00:00 | 6.323000 
949662 | 17030 | 11284035 | 30.730860 | 2008-09-09 00:00:00 | 6.335000 
949662 | 17030 | 11284035 | 16.290063 | 2008-12-10 00:00:00 | 111.045000 
950091 | 25885 | 11937183 | 14.629975 | 2012-03-23 00:00:00 | 12.337000 
950091 | 25885 | 11937183 | 14.599671 | 2012-04-23 00:00:00 | 12.140000 
950091 | 4325 | 11937183 | 11.082955 | 2012-01-27 00:00:00 | 6768.953000 
950091 | 4325 | 11937183 | 11.119163 | 2012-01-31 00:00:00 | 1.242000

来源

2012-07-12 Dan Snell

你看过[书籍中的EXCEPT和INTERSECT]（http://msdn.microsoft.com/en-us/library/ms188055.aspx）吗？ – 2012-07-12 17:19:02

@AaronBertrand我不认为EXCEPT和INTERSECT会在这里工作，因为已知AccountID在这两个查询中不匹配，除非OP可以删除该列 – 2012-07-12 17:22:30

实际上删除AccountID表是完全可能的。如果稍后需要，我总是可以使用UniqueAccountID作为参考。我会看看EXCEPT和INTERSECT。谢谢。 – 2012-07-12 17:27:35

我想你想要做一个完全外部联接，以找到那些不匹配：

select (case when c3.UniqueAccountId is not null and t2.UniqueAccountId is not null 
      then 'MATCH' 
      when c3.UniqueAccountId is not null then 'DEV-ONLY' 
      when t2.UniqueAccountId is not null then 'TEST-ONLY' 
      else 'Oops!' 
     end) as MatchType, 
     c3.UniqueAccountId as 'Dev3', t2.UniqueAccountId as 'Test2' 
from cteTest2 t2 full outer join 
    cteDev3 c3 
    on c3.UniqueAccountId = t2.UniqueAccountId and 
     c3.SecurityID = t2.SecurityID and 
     c3.BuyDate = t2.BuyDate and 
     c3.Shares = t2.Shares and 
     c3.BuyPrice = t2.BuyPrice

来源

2012-07-12 18:10:44

实际上，这对我所做的工作非常有效。谢谢。 – 2012-07-12 20:42:30

如何找到没有唯一ID的两个查询之间的差异

回答

相关问题