从每个员工的数据中获取前 1

Question

我有一个查询生成这样的数据

select c.Id, a.userId as Payroll, c.CurrentLocation, c.Longitude, c.Latitude, c.TextOtherLocation, c.CreationDate
    from PeopleDB.dbo.masterpeople a
    inner join PeopleDB.dbo.masterPeople b on a.manager = b.userid
    inner join PeopleTracking c on a.userId = c.payroll
    where a.Manager = '20090036'
ORDER BY Id DESC

Id  Payroll     CurrentLocation       Longitude     Latitude    TextOtherLocation   CreationDate
51  20180002    Field Work Location   107.5588565   -6.9077868                      6/13/2020 19:56
50  20180002    Field Work Location   107.5588565   -6.9077868                      6/14/2020 19:56
49  20190026    Office                107.5587918   -6.9077061                      6/15/2020 19:56
48  20190026    Field Work Location   107.5588565   -6.9077868                      6/16/2020 19:56
47  20190026    Office                107.5588565   -6.9077868                      6/17/2020 19:56

我怎样才能得到任何工资单的任何前 1 数据降序从上面的数据来看，我想要的是：

Id  Payroll     CurrentLocation       Longitude     Latitude    TextOtherLocation   CreationDate
51  20180002    Field Work Location   107.5588565   -6.9077868                      6/13/2020 19:56
49  20190026    Office                107.5587918   -6.9077061                      6/15/2020 19:56

感谢您的帮助。

Answer 1

使用row_number():

select t.*
from (select c.Id, a.userId as Payroll, c.CurrentLocation, c.Longitude, c.Latitude, c.TextOtherLocation, c.CreationDate,
             row_number() over (partition by a.userid order by c.CreatedDate desc) as seqnum
      from PeopleDB.dbo.masterpeople a join
           PeopleDB.dbo.masterPeople b
           on a.manager = b.userid join
           PeopleTracking c
           on a.userId = c.payroll
      where a.Manager = '20090036'
     ) t
where seqnum = 1
ORDER BY Id DESC;

请注意，a、b 和 c 确实是 table 别名的错误选择。这些应该是 table 缩写，分别类似于 mp1、mp2 和 pt。

Answer 2

我通常使用 row_number() 窗口函数将序数分配给这样的行。

select Id,Payroll,CurrentLocation,Longitude,Latitude, TextOtherLocation, CreationDate
from (
select c.Id, a.userId as Payroll, c.CurrentLocation, c.Longitude, c.Latitude, c.TextOtherLocation, c.CreationDate, row_number() over (partition by a.userId order by c.Id desc) as PayrollRowNum
    from PeopleDB.dbo.masterpeople a
    inner join PeopleDB.dbo.masterPeople b on a.manager = b.userid
    inner join PeopleTracking c on a.userId = c.payroll
    where a.Manager = '20090036') as PayrollData where PayrollRowNum = 1

这会为每一行分配原始编号，并为每一行重置编号 a.userId。然后根据每个 a.userId "group".

中的降序 c.Id 仅检索 "first" 一个

从每个员工的数据中获取前 1

Get Top 1 From Data For Every Employee

sql

sql-server

sql-server-2012

sql-server-2017