Question

我有数据表：

YYYY / MM / DD HH：MM：SS - 观看次数，点击次数等。
YYYY / MM / DD HH：MM：SS - ......

我希望将所有这些列加总为和值。喜欢： 1. YYYY / MM / DD - 观看总和，点击总和

我做了什么： 1.查询foreach列。对于观点;

select 
  cast([EventTime] as date) as 'Date',
  Count([Id]) as 'Views'
from [TelemetryData] 
where [DiscountId] = '8773fd1b-f0c0-41fd-b0a0-ab8238f227f5'
  and [EventName]     = 'DiscountView'
group by cast([EventTime] as date)
order by cast([EventTime] as date) asc

每日点击次数：

select 
  cast([EventTime] as date) as 'Date',
  Count([Id]) as 'Clicks'
from [TelemetryData] 
where [DiscountId] = '8773fd1b-f0c0-41fd-b0a0-ab8238f227f5'
  and [EventName] =     'DiscountClick'
group by cast([EventTime] as date)
order by cast([EventTime] as date) asc

我怎样才能将它们全部排成一行？

Answer 1

尝试检查和总结：

SELECT cast([EventTime] as date) as 'Date', SUM(case when [EventName]='DiscountClick' then 1 else 0 end) as 'Clicks', SUM(case when [EventName]='DiscountView' then 1 else 0 end) as 'Views'
from [TelemetryData] 
where [DiscountId] = '8773fd1b-f0c0-41fd-b0a0-ab8238f227f5'
and ([EventName] = 'DiscountView' or [EventName] = 'DiscountClick')
group by cast([EventTime] as date)
order by cast([EventTime] as date) asc

改变了什么：

我们首先选择了哪个列来过滤视图和点击。

然后我们使用SUM但是在EVENT = SOMETHING时使用CASE我们加1，否则为0。

通过您的论坛，您可以在一次查询中获得两天。

Answer 2

使用SUM和CASE ..

sum(case when [DiscountId] = '8773fd1b-f0c0-41fd-b0a0-ab8238f227f5'
  and [EventName] =     'DiscountClick' then noofclicks column else 0 end
) as clicks

sum(case when [DiscountId] = '8773fd1b-f0c0-41fd-b0a0-ab8238f227f5'
  and [EventName] =     'DiscountView' then view column else 0 end
) as views

以及其他......

from
table
group by cast([EventTime] as date)
order by cast([EventTime] as date)

SQL concat很少选择成一个大表

2 个答案: