根据问题和日期显示不同的总数答案

【问题标题】：Show distinct totals based on Issue & Date根据问题和日期显示不同的总数
【发布时间】：2020-02-06 11:22:58
【问题描述】：

我的查询有以下结果：

以上样本数据：

declare @tbl table (DateOfService date, IssueID int, IssueName varchar(100), AssignedConsultant varchar(30), Activity varchar(100), TimeSpent float, TimeInvoiced float)

insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2019-12-19',10295,'Bin Issue','Doug','Activity Detail 1…',10.25,8)
insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2019-12-19',10295,'Bin Issue','Doug','Activity Detail 2…',10.25,8)
insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2020-01-24',10295,'Bin Issue','Doug','Activity Detail 3…',5.5,4)
insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2020-01-24',10295,'Bin Issue','Doug','Activity Detail 4…',5.5,4)

select
*
from    @tbl

但是，出于保密原因，我无法分享我的查询，

如何将这些结果更改为如下所示：

从我的预期结果可以看出，我只需要每个日期花费的总时间和发票。

我遇到的问题是，这个结果被用于 Excel 中的数据透视，导致双值反映。

请帮忙？

【问题讨论】：

标签： tsql sql-server-2017

【解决方案1】：

我假设您有某种方法可以识别每一行，所以我在表变量中添加了 ID 列。我还将辅助行的值保留为 NULL，因为您不清楚您想要在那里的值。无论如何，这应该适用于表示的数据集：

declare @tbl table (ID int,DateOfService date, IssueID int, IssueName varchar(100), AssignedConsultant varchar(30), Activity varchar(100), TimeSpent float, TimeInvoiced float)

insert into @tbl (ID,DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values (1,'2019-12-19',10295,'Bin Issue','Doug','Activity Detail 1…',10.25,8)
insert into @tbl (ID,DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values (2,'2019-12-19',10295,'Bin Issue','Doug','Activity Detail 2…',10.25,8)
insert into @tbl (ID,DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values (3,'2020-01-24',10295,'Bin Issue','Doug','Activity Detail 3…',5.5,4)
insert into @tbl (ID,DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values (4,'2020-01-24',10295,'Bin Issue','Doug','Activity Detail 4…',5.5,4)


;WITH CTE AS (
SELECT MIN(ID) ID, DateOfService
FROM @tbl
GROUP BY DateOfService
)
SELECT T1.DateOfService,
        IssueID,
        IssueName,
        AssignedConsultant,
        Activity, 
        CASE WHEN T2.ID IS NULL THEN NULL ELSE T1.TimeSpent END TimeSpent,
        CASE WHEN T2.ID IS NULL THEN NULL ELSE T1.TimeInvoiced END TimeInvoiced
FROM @tbl T1
LEFT JOIN CTE T2 on T1.ID = T2.ID

【讨论】：

【解决方案2】：

我认为这行不通，但看起来对这些结果有效。

这是我想出的（不确定它在技术上是否正确），但它对我有用：

declare @tbl table (DateOfService date, IssueID int, IssueName varchar(100), AssignedConsultant varchar(30), Activity varchar(100), TimeSpent float, TimeInvoiced float)

insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2019-12-19',10295,'Bin Issue','Doug','Activity Detail 1…',10.25,8)
insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2019-12-19',10295,'Bin Issue','Doug','Activity Detail 2…',10.25,8)
insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2020-01-24',10295,'Bin Issue','Doug','Activity Detail 3…',5.5,4)
insert into @tbl (DateOfService,IssueID,IssueName,AssignedConsultant,Activity,TimeSpent,TimeInvoiced) values ('2020-01-24',10295,'Bin Issue','Doug','Activity Detail 4…',5.5,4)

select
    DateOfService
,   IssueID
,   IssueName
,   AssignedConsultant
,   Activity
,   case when lag(TimeSpent) over (order by IssueName) = TimeSpent then 0 else TimeSpent end            TimeSpent
,   case when lag(TimeInvoiced) over (order by IssueName) = TimeInvoiced then 0 else TimeInvoiced end   TimeInvoiced
from    @tbl

在db<>fiddle查看结果

【讨论】：