I saw that using GROUP BY CUBE()
would add a Total row to the bottom of your pivot, and I have been trying to get my syntax accurate so that I can do such. Well I have my syntax running, but I am not getting the Total at the bottom of my result set. Why is that?
select *
FROM
(
select case
WHEN a.state LIKE 'CA' THEN 'California'
WHEN a.state LIKE 'WA' THEN 'Washington'
else a.state
end As [Full State],
SaleID As [Sales By State],
CONVERT(VARCHAR(20), dt.CumulativeWeek) AS Week
FROM retailsales.Store1 a
INNER JOIN retailsales.customCalendar dt
ON a.orderDate = dt.orderDate
WHERE a.orderDate IS NOT NULL
GROUP BY CUBE (SaleID, state, dt.CumulativeWeek)
) src
pivot
(
COUNT([Sales By State])
For Week IN ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12],[13])
) piv
I tried moving the CUBE()
statement outside of my pivot but am getting a compile error here is updated code
select *
FROM
(
select case
WHEN a.state LIKE 'CA' THEN 'California'
WHEN a.state LIKE 'WA' THEN 'Washington'
else a.state
end As [Full State],
SaleID As [Sales By State],
CONVERT(VARCHAR(20), dt.CumulativeWeek) AS Week
FROM retailsales.Store1 a
INNER JOIN retailsales.customCalendar dt
ON a.orderDate = dt.orderDate
WHERE a.orderDate IS NOT NULL
GROUP BY CUBE (SaleID, state, dt.CumulativeWeek)
) src
pivot
(
COUNT([Sales By State])
For Week IN ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12],[13])
) piv
GROUP BY CUBE (Full State)
And that produces this compile error
Msg 8120, Level 16, State 1, Line 1
Column 'piv.1' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause.