如果标题不够清楚,我很抱歉,但在Qlikview中可以这样做吗?
Variable Status Date Duration (Hours)
A StatusA 9/10/2017 18:30:00.00 4
A StatusB 9/10/2017 23:30:00.00 5
B StatusA 9/10/2017 10:00:00.00 8
B StatusB 9/10/2017 21:45:00.00 9
以及我希望它如何处理。
Variable Status Date Duration (Hours) FinishDate
A StatusA 9/10/2017 18:30:00.00 4 9/10/2017 22:30:00.00
A StatusB 9/10/2017 23:30:00.00 0.5 9/10/2017 23:59:59.59
A StatusB 9/11/2017 0:00:00.00 4.5 9/11/2017 3:30:00.00
B StatusA 9/10/2017 10:00:00.00 8 9/10/2017 18:00:00.00
B StatusB 9/10/2017 21:45:00.00 2.25 9/10/2017 23:59:59.59
B StatusB 9/11/2017 0:00:00.00 6.75 9/11/2017 6:15:00.00
我知道这可以通过常驻表实现,但如何检查
变量的
运行时间超过一天,然后创建另一行以分隔到第二天的持续时间。这是因为我每周都有一个过滤器,所以如果最后一天(或周六)有
变量
如果超过持续时间,数据将不准确。
结果数据将需要获得每个
变量
脚本:
T1:
LOAD *, timestamp(Timestamp+[Duration Hours]/24) as FinishDate;
LOAD *, timestamp(Timestamp#(left(Date,19),'DD-MM-YYYY hh:mm:ss')) as Timestamp;
LOAD
[EquipmentID] AS [Equipment ID],
[TransactionDate] AS [Date],
[LotID] AS [Lot ID],
[Status] AS [Status],
[DurationHours] AS [Duration Hours];
SQL
SELECT *
FROM [SQL_SourceDB].[dbo].[SourceTable]
WHERE [TransactionDate] >= '2016-01-01 00:00:00.000' AND [TransactionDate] <= '2016-01-31 00:00:00.000';
Left Join // add a split-flag where needed
LOAD Distinct Timestamp,FinishDate,fabs(Date(left(FinishDate,10))-Date(left(Timestamp,10)) >=1) as SplitFlag
Resident T1;
T2: // load first part (current day) of split-flag=1
LOAD
[Equipment ID],
[Timestamp] AS [Date],
[Lot ID],
[Status],
round((DayEnd(Timestamp)-Timestamp)*24,0.1) AS [Duration Hours]
Resident T1
Where SplitFlag=1;
Concatenate // load second part (next day) where split-flag=1
LOAD
[Equipment ID],
daystart(FinishDate) AS [Date],
[Lot ID],
[Status],
round((FinishDate-daystart(FinishDate))*24,0.1) AS [Duration Hours]
Resident T1
Where SplitFlag=1;
Concatenate // add the rest of the data (split-flag=0)
LOAD * Resident T1 Where SplitFlag=0;
DROP Table T1;