Suppose I have the following.
select
case
when fcompany = \'Acme\' then \'Red\'
when fcompany = \'Acme Rockets\' then \'Blue\'
else \'Green\'
en
Try this:
SELECT b.Color,
sum(fann_sales)
FROM (
SELECT case
when fcompany = 'Acme' then 'Red'
when fcompany = 'Acme Rockets' then 'Blue'
else 'Green'
end
Color,
fann_sales
FROM slcdpm
) a RIGHT JOIN
(
SELECT 'Red' AS Color
UNION ALL
SELECT 'Blue' AS Color
UNION ALL
SELECT 'Green' AS Color
) b
ON a.Color = b.Color
GROUP BY b.Color
If all colors must be present, why not present them as columns?
SELECT
(SELECT sum(fann_sales) FROM slcdpm WHERE fcompany = 'Acme') AS RedSum,
(SELECT sum(fann_sales) FROM slcdpm WHERE fcompany = 'Acme Rockets') AS BlueSum,
(SELECT sum(fann_sales) FROM slcdpm WHERE fcompany <> 'Acme' AND fcompany <> 'Acme Rockets') AS GreenSum
Otherwise, just go with @JohnK813 answer.
Move the GROUP into a conditional SUM with more columns?
select
sum(CASE WHEN fcompany = 'Acme'
THEN fann_sales ELSE 0 END) AS redsales,
sum(CASE WHEN fcompany = 'Acme Rockets'
THEN fann_sales ELSE 0 END) AS bluesales
sum(CASE WHEN fcompany NOT IN ('Acme Rockets', 'Acme')
THEN fann_sales ELSE 0 END) AS greensales
FROM
slcdpm
One pass over the table for this. A UNION ALL or subquery approach (in other answers) will touch the table once per clause = somewhat slower.
Yes, Union All may be your best bet.
SELECT 'red' AS Color, sum(fann_sales) FROM slcdpm WHERE fcompany = 'Acme' GROUP BY fcompany
UNION ALL
SELECT 'blue' AS Color, sum(fann_sales) FROM slcdpm WHERE fcompany = 'Acme Rockets' GROUP BY fcompany
UNION ALL
SELECT 'green' AS Color, sum(fann_sales) FROM slcdpm WHERE fcompany <> 'Acme' AND fcompany <> 'Acme Rockets' GROUP BY fcompany