SQL SELECT multiple count in one query - sql

Question

I have these two tables:
and I want get this result:
How can I achieve this by using only one query?
I tried with join and count and group by but I cannot get it right.
I tried this already, but I cannot get it to work properly.
SELECT
coupon.*,
couponUsers.returned AS COUPON_TOTAL_USERS,
couponUses.returned AS COUPON_TOTAL_USES
FROM
coupon,
(SELECT
coupon.COUPON_CODE,
COUNT(redeemed.REDEEMED_USER) AS returned
FROM
coupon
JOIN
redeemed ON coupon.COUPON_CODE = redeemed.REDEEMED_CODE
GROUP BY
redeemed.REDEEMED_USER) couponUsers,
(SELECT
coupon.COUPON_CODE,
COUNT(redeemed.REDEEMED_CODE) AS returned
FROM
coupon
JOIN
redeemed ON coupon.COUPON_CODE = redeemed.REDEEMED_CODE
GROUP BY
redeemed.REDEEMED_CODE) couponUses
WHERE
coupon.COUPON_CODE = couponUsers.COUPON_CODE
AND coupon.COUPON_CODE = couponUses.COUPON_CODE
GROUP BY
coupon.COUPON_CODE
ORDER BY
coupon.COUPON_ID ASC
This is the build schema if you want to try it yourself in SQL fiddle or something like that..
CREATE TABLE IF NOT EXISTS `coupon`
(
`COUPON_ID` int(11) NOT NULL,
`COUPON_CODE` varchar(32) NOT NULL
) DEFAULT CHARSET=utf8;
INSERT INTO `coupon` (`COUPON_ID`, `COUPON_CODE`) VALUES
(1, "AAAAA"),
(2, "BBBBB"),
(3, "CCCCC"),
(4, "DDDDD"),
(5, "EEEEE");
CREATE TABLE IF NOT EXISTS `redeemed` (
`REDEEMED_ID` int(11) NOT NULL,
`REDEEMED_USER` varchar(32) NOT NULL,
`REDEEMED_CODE` varchar(32) NOT NULL
) DEFAULT CHARSET=utf8;
INSERT INTO `redeemed` (`REDEEMED_ID`, `REDEEMED_USER`, `REDEEMED_CODE`) VALUES
(1, "TOM", "AAAAA"),
(2, "PAULA", "BBBBB"),
(3, "TOBI", "CCCCC"),
(4, "JANA", "DDDDD"),
(5, "INGO", "EEEEE"),
(6, "TOM", "AAAAA"),
(7, "PETER", "EEEEE"),
(8, "JIM", "DDDDD"),
(9, "SARA", "AAAAA"),
(10, "TOBI", "CCCCC"),
(11, "PAULA", "AAAAA"),
(12, "TOM", "AAAAA"),
(13, "PAULA", "BBBBB"),
(14, "JIM", "DDDDD"),
(15, "JANA", "DDDDD");
i am trying this already a couple hours..
its time for some help ^^

Answer 1

You should be able to generate the wanted counts in a single table query:
select
redeemed_code
, count(*) as tot_uses
, count(distinct redeemed_user) as tot_users
from redeemed
group by redeemed_code
You can join this to the coupon table for final output, for example with a left join you would get all coupons listed in coupon even if none have been redeemed yet.
select
c.coupon_id
, c.coupon_code
, coalesce(d.tot_uses,0) as tot_uses
, coalesce(d.tot_users,0) as tot_users
from coupon as c
left join (
select
redeemed_code
, count(*) as tot_uses
, count(distinct redeemed_user) as tot_users
from redeemed
group by redeemed_code
) as d on c.coupon_code = d.redeemed_code
coupon_id
coupon_code
tot_uses
tot_users
1
AAAAA
5
3
2
BBBBB
2
1
3
CCCCC
2
1
4
DDDDD
4
2
5
EEEEE
2
2
db<>fiddle here

Answer 2

This is for PostgreSQL, but this query should work for you. You can do the GROUP BY on the redeemed table and join that with coupon to get the COUPON_ID. \
psql=# WITH redeemed_query AS (
SELECT
"REDEEMED_CODE",
COUNT(*) AS "TOTAL_USES",
COUNT(DISTINCT("REDEEMED_USER")) AS "TOTAL_USERS"
from redeemed GROUP BY "REDEEMED_CODE"
)
SELECT * FROM redeemed_query
INNER JOIN
coupon ON redeemed_query."REDEEMED_CODE" = coupon."COUPON_CODE";
REDEEMED_CODE | TOTAL_USES | TOTAL_USERS | COUPON_ID | COUPON_CODE
---------------+------------+-------------+-----------+-------------
AAAAA | 5 | 3 | 1 | AAAAA
BBBBB | 2 | 1 | 2 | BBBBB
CCCCC | 2 | 1 | 3 | CCCCC
DDDDD | 4 | 2 | 4 | DDDDD
EEEEE | 2 | 2 | 5 | EEEEE
(5 rows)