Reputation: 331
Say I have a table like this in ms sql 2008:
+------+--------+---------+
| year | JAN | FEB |
+------+--------+---------+
| 2016 | 5K2 | 5K2 |
| 2016 | 5K2 | 5K2 |
| 2016 | 5K2 | 5K2 |
| 2016 | 8Z | 8Z |
| 2016 | R5205 | R5205 |
| 2016 | 5K2 | 5K2 |
| 2016 | 5K2 | 5K2 |
| 2016 | NULL | NULL |
| 2016 | TE | NULL |
| 2016 | TE | NULL |
| 2016 | 8Z | 8Z |
+------+--------+---------+
And I want to get a count for each column, something like this
+------+--------+---------+
| opt | JAN_cnt| FEB_cnt |
+------+--------+---------+
| 5K2 | 5 | 4 |
| 8Z | 2 | 2 |
| R5205| 1 | 1 |
| TE | 2 | 0 |
| NULL | 1 | 4 |
+------+--------+---------+
First, can this be done? Second, how? I have searched, but cant find exactly what I am looking for.
Upvotes: 1
Views: 583
Reputation: 40471
I think the simplest way is to use UNION ALL
with conditional aggregation using CASE EXPRESSION
:
SELECT s.opt,
COUNT(CASE WHEN s.ind_from = 1 THEN 1 END) as jan_cnt,
COUNT(CASE WHEN s.ind_from = 2 THEN 1 END) as feb_cnt
FROM (
SELECT t1.jan as opt,1 as ind_from FROM YourTable t1
UNION ALL
SELECT t2.feb,2 FROM YourTable t2) s
GROUP BY s.opt
Upvotes: 1
Reputation: 520878
SELECT COALESCE(t1.JAN, t2.FEB), t1.JAN_cnt, t2.FEB_cnt
FROM
(
SELECT JAN, COUNT(*) AS JAN_cnt
FROM yourTable
GROUP BY JAN
) t1
FULL OUTER JOIN
(
SELECT FEB, COUNT(*) AS FEB_cnt
FROM yourTable
GROUP BY FEB
) t2
ON t1.JAN = t2.FEB
Upvotes: 0
Reputation: 1269445
I would advise putting the values into a different format:
You can do this as:
select opt, mon, count(*) as cnt
from ((select jan as opt, 'jan' as mon from t) union all
(select feb as opt, 'feb' as mon from t)
) o
group by opt, mon;
It is easy enough to switch this to your format:
select opt, sum(jan) as jan, sum(feb) as feb
from ((select jan as opt, 1 as jan, 0 as feb from t) union all
(select feb as opt, 0, 1, from t)
) o
group by opt;
I just prefer the first format. It is easier to generalize to more columns.
Upvotes: 0