Unfortunately I can't use the lag
function due to using SQL Server 2008.
I need to find machine based difference between two values at different times.
I have a table like that:
ID | Time | Value | Machine
------ ----------- --------- -----------
7 | 12:00 | 154400 | 20
6 | 12:00 | 77840 | 17
5 | 11:00 | 10480 | 11
4 | 11:00 | 13214 | 14
3 | 11:00 | 76340 | 17
2 | 11:00 | 15410 | 20
1 | 10:00 | 75000 | 17
I need to find a difference (76340-75000 = 1340) between id 1 and 3.
Tried these before:
select
t1.Machine, t1.[ID], t2.[ID], t1.Value, t2.Value,
t2.Value - t1.Value as difference
from
hava t1
cross apply
hava t2
where
t1.Machine = t2.Machine
and t1.ID <> t2.ID
and t1.Machine = 17
but it compares with t1.id with the whole table.
I can't use lag
function - thank you very much for your help.
CodePudding user response:
WITH CTE(ID , Time, Value, Machine)AS
(
SELECT 7,'12:00',154400,20 UNION ALL
SELECT 6,'12:00',77840,17 UNION ALL
SELECT 5,'11:00',10480,11 UNION ALL
SELECT 4,'11:00',13214,14 UNION ALL
SELECT 3,'11:00',76340,17 UNION ALL
SELECT 2,'11:00',15410,20 UNION ALL
SELECT 1,'10:00',75000,17
),
CTE2 AS
(
SELECT C.ID,C.TIME,C.Value,C.Machine,
ROW_NUMBER()OVER(PARTITION BY C.MACHINE ORDER BY C.TIME ASC)XCOL
FROM CTE AS C
)
SELECT X.ID,X.Time,X.Value, X2.Value, X.Machine
FROM CTE2 AS X
LEFT JOIN CTE2 AS X2 ON X.Machine=X2.Machine AND X.XCOL=X2.XCOL-1
WHERE X.Machine=17