SQL Server比较仅在一个表中匹配的数据量

Question

我有两个表：CustomerInformation

CustomerName CustomerAddress CustomerID LocationID   BillDate
CITY - 1    500 N ST    47672001    29890   2012-07-20 00:00:00.000 0
CITY - 1    500 N ST    47672001    29890   2012-07-20 00:00:00.000 6890
CITY - 1    500 N ST    47672001    29890   2012-08-17 00:00:00.000 0
CITY - 9    510 N ST    47643241    29890   2012-08-17 00:00:00.000 5460
CITY - 4213 500 S ST    43422001    29890   2012-09-17 00:00:00.000 0
CITY - 5    100 N ST    23272001    29890   2012-09-17 00:00:00.000 4940
CITY - 3    010 N ST    43323001    29890   2012-10-19 00:00:00.000 0
CITY - 78   310 N ST    12222001    29890   2012-10-19 00:00:00.000 5370

和CustomerMeters具有三列：ID，名称，地址

这两个表之间的连接是：CustomerAddress，因此我可以基于Address将两个表连接起来：

SELECT * FROM CustomerInformation 
JOIN CustomerMeters 
ON CustomerAddress  = Address 

现在，问题是我有这么多记录（CustomerInformation中的20000以上），难道我没有列出两个表中匹配的记录数，以及CustomerInformation表中只有多少条记录？

谢谢。

Answer 1

联接产生的记录数：

SELECT COUNT(*) 
FROM CustomerInformation 
JOIN CustomerMeters 
  ON CustomerAddress = Address

CustomerInformation表中专有的记录数：

SELECT COUNT(*) 
FROM CustomerInformation AS CI -- Records in CustomerInformation
WHERE NOT EXISTS(SELECT *      -- that are not in CustomerMeters
                 FROM CustomerMeters AS CM
                 WHERE CM.Address = CI.CustomerAddress)

Answer 2

随后的查询将为您提供CustomerInformation表中所有记录的列表和一个标志列MATCH ，如果CustomerMeters表中存在核心响应记录，则该标志列将包含1，否则为0。

SELECT  CI.ID
        ,Ci.Name
        ,CI.CustomerAddress 
        ,CASE WHEN CM.Address IS NULL THEN 0 ELSE 1 END AS MATCH
FROM    CustomerInformation CI
LEFT OUTER JOIN
        CustomerMeters CM 
ON      CM.Address = CI.CustomerAddress 

Answer 3

select
    C.grp, count(*)
from CustomerInformation as ci
    left outer join CustomerMeters as cm on cm.CustomerAddress = ci.Address
    outer apply (
        select
            case
                when cm.ID is not null then 'Number of records in both tabless'
                else 'not in CustomerMeters'
            end as grp
    ) as C
group by C.grp

要么

--number of records in both tables
select count(*)
from CustomerInformation as ci
where ci.Address in (select cm.CustomerAddress from CustomerMeters as cm)

--number of records in CustomerInformation which are not in CustomerMeters
select count(*)
from CustomerInformation as ci
where ci.Address not in (select cm.CustomerAddress from CustomerMeters as cm)