如何從日期列中剝離時間並使每個日期的用戶數唯一

Question

我有一個事件表，其中包含太多列。 我想查詢3列，DateTimeResolved，團隊和個人。 首先，我想將DateTimeResolved列中的時間與日期分開。 然后我想知道自2017年4月以來，每個團隊有多少個人（唯一身份）解決了事件。

能做到嗎？

我有以下查詢工作，但顯然它們需要調整，因為它們沒有給我我想要的最終結果。

select stat_datetimeresolved, resolvedbyteam, resolvedby 
from [dbo].[Incident]
where Stat_DateTimeResolved is not null
and Stat_DateTimeResolved >= '2017-04-01 00:00:00'
order by Stat_DateTimeResolved asc


select CAST(stat_datetimeresolved as date),
COUNT(resolvedby)
from [dbo].[Incident]
group by CAST(Stat_DateTimeResolved as DATE)
order by CAST(stat_datetimeresolved as DATE) asc

Answer 1

你很親密 您可以將DISTINCT關鍵字添加到COUNT中，以按天獲取唯一身份的團隊成員數。

SELECT CAST(stat_datetimeresolved as date) as DateResolved
, ResolvedByTeam
, COUNT(DISTINCT resolvedby) as ResolvedByCnt
FROM [dbo].[Incident]
 WHERE Stat_DateTimeResolved >= '20170401'
GROUP BY CAST(Stat_DateTimeResolved as DATE), ResolvedByTeam
ORDER BY CAST(stat_datetimeresolved as DATE) asc, ResolvedByTeam

Answer 2

您在此處的問題中有多個不兼容的請求。 我認為您所追求的實際上是：

自2017年4月起，每個團隊每天有多少人制定決議？

為此，您只需要返回一個唯一的列表，其中包含解決日期，個人所屬的團隊以及實際做出解決的團隊，然后進行匯總：

with r as
(
    select distinct cast(stat_datetimeresolved as date) as DateResolved
                   ,ResolvedByTeam
                   ,ResolvedBy
    from [dbo].[Incident]
    where Stat_DateTimeResolved >= '20170401'  -- NULL check is redundant if you are filtering for a value
)
select DateResolved
      ,ResolvedByTeam
      ,count(ResolvedBy) as NumIndividualsMakingResolutions
from r
group by DateResolved
        ,ResolvedByTeam
order by DateResolved
        ,ResolvedByTeam;