SQL Server-将5个查询合并到一个表中

您好，我必须创建一份报告，说明药店的出境使用量 有4种类型的出站类型：SMS、呼叫、电子邮件和P

我想得到每种类型的总数以及每种药房的总数

以下是我目前的疑问

select PharmacyID, count(*) as total
from OutboundCallData
where datepart(year, DateReceived) = 2014 and datepart(month, datereceived) = 1
        and (NotificationMode = 'sms' or NotificationMode = 'call' or NotificationMode = 'email' or NotificationMode = 'p')
group by PharmacyID

select PharmacyID, count(*) as sms
from OutboundCallData
where datepart(year, DateReceived) = 2014 and datepart(month, datereceived) = 1
        and NotificationMode = 'sms'
group by PharmacyID

select PharmacyID, count(*) as call
from OutboundCallData
where datepart(year, DateReceived) = 2014 and datepart(month, datereceived) = 1
        and NotificationMode = 'call'
group by PharmacyID


select PharmacyID, count(*) as email
from OutboundCallData
where datepart(year, DateReceived) = 2014 and datepart(month, datereceived) = 1
        and NotificationMode = 'email'
group by PharmacyID


select PharmacyID, count(*) as p
from OutboundCallData
where datepart(year, DateReceived) = 2014 and datepart(month, datereceived) = 1
        and NotificationMode = 'p'
group by PharmacyID

我想把所有这些合并成一张这样的表

PharmacyID| SMS | Call | Email | P | Total
-----------------------------------------
999000001 |  3  |  4   |   5   | 6 |  18
999000002 |  12 |  0   |   14  | 8 |  34

---

显示0或null就可以了。

一个简单的交叉制表符可以：

select PharmacyID,
    count(*) as total,
    count(case when NotificationMode = 'sms' then 1 else null end) as sms,
    count(case when NotificationMode = 'email' then 1 else null end) as email,
    count(case when NotificationMode = 'call' then 1 else null end) as call,
    count(case when NotificationMode = 'p' then 1 else null end) as p
from OutboundCallData
where datepart(year, DateReceived) = 2014 and datepart(month, datereceived) = 1
        and (NotificationMode = 'sms' or NotificationMode = 'call' or NotificationMode = 'email' or NotificationMode = 'p')
group by PharmacyID

或者，您可以使用

轴

：

SELECT PharmacyID, [sms], [email], [call], [p]
FROM
    (SELECT PharmacyID, NotificationMode 
        from OutboundCallData) AS outboundData
    PIVOT
    (
       count(NotificationMode )
       FOR NotificationMode IN ([sms], [email], [call], [p])
    ) AS P

---

下面是一个使用PIVOT关键字的示例

SELECT PharmacyID, [sms], [email], [call], [p]
FROM
    (SELECT PharmacyID, NotificationMode 
        from OutboundCallData
        where datepart(year, DateReceived) = 2014 and datepart(month, datereceived) = 1) o
    PIVOT
    (
       count(NotificationMode )
       FOR NotificationMode IN ([sms], [email], [call], [p])
    ) AS [PivotTable]

---

这种技术称为交叉表。我个人更喜欢这个而不是一个支点。我发现它们更容易阅读，而且在性能上也略胜一筹d感谢您的帮助，我在关键字“PIVOT”附近遇到一个错误，语法不正确。这是由于在末尾使用了

PIVOT

作为表名。我现在已经补救了。