Question

我有以下查询：

select distinct 
(SELECT  COUNT (TaskID) FROM Task, Employee where SupervisorID = '752A0F4D-9905-4D55-87F1-AFA19245C206' and task.UserID = Employee.UserID) as Total,
(SELECT  COUNT (TaskID) FROM Task, Employee where SupervisorID = '752A0F4D-9905-4D55-87F1-AFA19245C206' and task.UserID = Employee.UserID and DateDue <= DATEADD(day,-1, GETDATE()) )as TotalOverdue,
(SELECT  COUNT (TaskID) FROM Task, Employee where SupervisorID = '752A0F4D-9905-4D55-87F1-AFA19245C206' and task.UserID = Employee.UserID and DateDue >= DATEADD(day,-1, GETDATE()) )  as Totaldue
from Task

它给了我以下结果：

Total | TotalOverdue | TotalDue
5     | 4            | 1

但是，我希望它以这种方式形成：

Total: 5
TotalOverdue: 4
TotalDue: 1

我尝试过使用pivot功能，但是无法让它正常工作。有什么想法吗？

Answer 1

我认为你可以通过条件聚合做你想做的事情：

select count(*) as Total,
       sum(case when DateDue <= DATEADD(day, -1, GETDATE()) then 1 else 0
           end) as TotalOverdue,
       sum(case when DateDue >=  DATEADD(day, -1, GETDATE()) then 1 else 0
           end) as TotalDue
from Task join
     Employee
     on task.UserID = Employee.UserID
where SupervisorID = '752A0F4D-9905-4D55-87F1-AFA19245C206';

如果你想要所有的主管，

select SupervisorID, count(*) as Total,
       sum(case when DateDue <= DATEADD(day, -1, GETDATE()) then 1 else 0
           end) as TotalOverdue,
       sum(case when DateDue >=  DATEADD(day, -1, GETDATE()) then 1 else 0
           end) as TotalDue
from Task join
     Employee
     on task.UserID = Employee.UserID;

编辑：

你真的想要取消数据的转移。这是一种方法：

select n.which,
       (case when n.which = 'Total'
             then count(*)
             when n.which = 'TotalOverdue'
             then sum(case when DateDue <= DATEADD(day, -1, GETDATE()) then 1 else 0
                                 end)
             else sum(case when DateDue >=  DATEADD(day, -1, GETDATE()) then 1 else 0
                      end)
        end) as value
from Task join
     Employee
     on task.UserID = Employee.UserID cross join
     (select 'Total' as which union all 'TotalOverdue' union all 'TotalDue') n
group by n.which;

如果您想要一个字符串列，请使用concat()。

具有多个计数的Sql枢轴

1 个答案: