我需要使用 CASE 语句合并两个单独的查询答案

【问题标题】：I need to merge two separate queries using CASE statement我需要使用 CASE 语句合并两个单独的查询
【发布时间】：2015-11-23 21:13:07
【问题描述】：

所以我需要使用 CASE 语句合并这两个单独的查询。我的总体目标是“我希望找到所有直播和免费直播的地点，而不是演示。这些地点必须位于美国或加拿大。找出这些地点有多少客户填写了他们的国家/地区不是美国或加拿大，而不是他们有多少客户。”

所以我把两个查询都写出来了，它们都是分开工作的。

这是第一个返回非美国客户的。

SELECT 
    COUNT(1) AS "Not in US or Canada verse Total Customers" 
FROM 
    dbo.Spa (nolock)
LEFT JOIN 
    dbo.Customer ON dbo.Customer.SpaID = dbo.Spa.ID
WHERE 
    dbo.Spa.TimeZoneID IN (6, 9, 12, 18) 
    AND dbo.Spa.IsDeleted = 0
    AND dbo.Spa.StatusID IN (3, 4)
    AND dbo.Customer.CountryID != 1
    AND dbo.Customer.countryID != 2;

这是返回所有客户的第二个。

SELECT 
    COUNT(1) AS "Total Customers" 
FROM 
    dbo.Spa (nolock)
LEFT JOIN 
    dbo.Customer ON dbo.Customer.SpaID = dbo.Spa.ID
WHERE 
    dbo.Spa.TimeZoneID IN (6, 9, 12, 18) 
    AND dbo.Spa.IsDeleted = 0
    AND dbo.Spa.StatusID IN (3, 4);

谁能帮我用case语句把这些放在一起，它会使用CASE语句在一个表中显示不在美国和加拿大的客户和总客户。

谢谢！

【问题讨论】：

标签： sql sql-server

【解决方案1】：

如果我理解你的问题，我认为这就是你要找的：

SELECT      COUNT(1) AS "All customers",
            SUM(CASE WHEN (dbo.Customer.CountryID != 1 AND dbo.Customer.countryID != 2)
                THEN 1 ELSE 0 END) AS "Not in US or Canada verse Total Customers"
FROM        dbo.Spa (nolock)
LEFT JOIN   dbo.Customer    ON  dbo.Customer.SpaID = dbo.Spa.ID
WHERE       dbo.Spa.TimeZoneID IN(6,9,12,18) 
AND         dbo.Spa.IsDeleted= 0
AND         dbo.Spa.StatusID IN (3,4)

【讨论】：

【解决方案2】：

您可以使用 case 表达式来做到这一点。

SELECT sum(case when dbo.Customer.CountryID <> 1 AND dbo.Customer.countryID <> 2 then 1 end) as NotInUS
    , COUNT(1) AS "Not in US or Canada verse Total Customers" 
FROM dbo.Spa (nolock)
LEFT JOIN dbo.Customer ON dbo.Customer.SpaID = dbo.Spa.ID
WHERE dbo.Spa.TimeZoneID IN(6,9,12,18) 
AND dbo.Spa.IsDeleted= 0
AND dbo.Spa.StatusID IN (3,4)

此外，您可能应该改掉将 NOLOCK 溅到各处的习惯。它有一些非常严重且经常被误解的副作用。 http://blogs.sqlsentry.com/aaronbertrand/bad-habits-nolock-everywhere/

【讨论】：

【解决方案3】：

您可以让每个查询成为子选择：

SELECT (query1) as '', (query2) as ''

【讨论】：

【解决方案4】：

在您的第一个查询中，您使用左连接，然后过滤 CountryID，这将消除没有匹配国家/地区的 Spa 行。实际上可能没有，但您应该知道，由于该更改，您可能会在总计中获得不同的计数。如果每个“Spa”（？）不能有超过一个客户，那么加入该表（在第二个查询中）甚至不会完成任何事情，但很难猜测您是否打算这样做。

select
    count(case when c.CountryID not in (1, 2) then 1 else null end),
    count(*)
from
    dbo.Spa as s
    left outer join dbo.Customer as c
        on c.SpaID = s.SpaID
where
        s.TimeZoneID in (6, 9, 12, 18)
    and s.IsDeleted = 0
    and s.StatusID in (3, 4)

【讨论】：

【解决方案5】：

这是一个如何在 COUNT() 中使用 CASE 语句来仅计算指定组的示例：

    WITH
    /*
    Get a list of all customers and their country to use as the main data set as an example
    */
    mainSet AS
        (
        SELECT
            contact.FirstName,
            address.city,
            stateProvince.countryRegionCode 
        FROM person.address AS address
        INNER JOIN sales.customerAddress AS customerAddress ON
            address.addressID = customerAddress.addressID
        INNER JOIN sales.customer AS customer ON
            customerAddress.customerID = customer.customerID
        INNER JOIN person.contact AS contact ON 
            customer.customerID = contact.contactID
        INNER JOIN person.stateProvince AS stateProvince ON
            address.stateProvinceID = stateProvince.stateProvinceID
        )

    SELECT
        --------------------------------------------------Count all customers using COUNT()
        COUNT(countryRegionCode) AS 'All Customers Count',

        --------------------------------------------------Count all customers using COUNT() with a CASE statement
        COUNT(CASE WHEN countryRegionCode NOT IN('US','CA') THEN countryRegionCode END) AS 'Customers not in US/CA'
    FROM mainSet

【讨论】：