Reputation: 627
Sorry bad topic… I need to calculate a running total but need to reset the total on a condition (when expected reached = 0). I have this table:
Date, Registrations, Expected Registrations, Expected reached
2020-03-01, 5, 4,1
2020-03-02, 7, 5,1
2020-03-03, 8, 6,1
2020-03-04, 2, 5,0
2020-03-05, 5, 4,1
2020-03-06, 7, 5,1
2020-03-07, 8, 6,1
2020-03-08, 2, 5,0
Expected result with running total - the condition is that while “Expected Reached” <> 0 running total should be calculated. If “Expected Reached” = 0 the running total should start over from 0:
Date, Registrations, Expected Registrations, Expected Reached, Running Total
2020-03-01, 5, 4,1, 1
2020-03-02, 7, 5,1, 2
2020-03-03, 8, 6,1, 3
2020-03-04, 2, 5,0, 0
2020-03-05, 5, 4,1, 1
2020-03-06, 7, 5,1, 2
2020-03-07, 8, 6,1, 3
2020-03-08, 2, 5,0, 0
I don't know how to partition my window function to do this. Maybe I have to create an intermediate calculation before but Im unsure. Any suggestions?
Edit: edit2: removed my "on the fly question".
Upvotes: 2
Views: 2978
Reputation: 172944
Below is for BigQuery Standard SQL
#standardSQL
SELECT * EXCEPT(grp),
SUM(Expected_reached) OVER(PARTITION BY grp ORDER BY `date`) Running_Total
FROM (
SELECT *, COUNTIF(Expected_reached = 0) OVER(ORDER BY `date`) grp
FROM `project.dataset.table`
)
Upvotes: 2
Reputation: 1269445
Use a cumulative sum to define "islands". Then use row_number()
:
select t.*,
(case when expected_reached > 0
then row_number() over (partition by grp, (expected_reached = 0)
order by date
)
else 0
end) as running_total
from (select t.*,
countif(expected_reached = 0) over (order by date) as grp
from t
) t;
Upvotes: 0