Incremental sales Output - sql

Question

I have a table like Bikes
name mfgdate qtysold
bajaj 1/1/2016 12:00:00 AM 48
bajaj 1/1/2017 12:00:00 AM 49
bajaj 1/1/2018 12:00:00 AM 50
pulsar 1/1/2016 12:00:00 AM 300
pulsar 1/1/2017 12:00:00 AM 250
pulsar 1/1/2018 12:00:00 AM 200
yamaha 1/1/2016 12:00:00 AM 90
yamaha 1/1/2017 12:00:00 AM 180
yamaha 1/1/2018 12:00:00 AM 100
From the above table i need a select Query to get the output of the increased sales
like
eg: output
name
Bajaj
because Bajaj has the increased sales. Yamaha also have a increased sales but it decreased in 2018..
Thanks in advance..

Answer 1

Based on OP's comments, it is required to find only those Bike names, which has never seen decrease in sales.
Following solution will work on All MySQL Versions (especially < 8.0)
Using a Correlated subquery, we find previous qtysold value for a bike. If not found, using Ifnull() function, consider it as zero
Calculate change (qtysold - previous qtysold)
Now, using this as Derived table, we can get all those Distinct bike name(s), where change has always been > 0
You can try the following (check SQL Fiddle):
SELECT t3.name
FROM
(
SELECT t1.name,
(t1.qtysold - IFNULL(
(SELECT t2.qtysold
FROM Bikes as t2
WHERE t2.name = t1.name
AND t2.mfgdate < t1.mfgdate
ORDER BY t2.mfgdate DESC LIMIT 1)
, 0
)) AS sales_change
FROM Bikes as t1
) AS t3
GROUP BY t3.name
HAVING MIN(t3.sales_change) > 0

Answer 2

You can use a SELF JOIN to get desired result. Inner query will return the names of all Bikes which have a dip in sales. You can use NOT IN clause then to find the left-out result entries.
SELECT DISTINCT b.name
FROM Bikes b
WHERE
b.name NOT IN (
SELECT DISTINCT b1.name
FROM
Bikes b1
INNER JOIN
Bikes b2
ON b1.name = b2.name
AND b1.mfgdate < b2.mfgdate
AND b1.qtysold > b2.qtysold
);

Answer 3

You can use window function :
select t.*,
sum(case when exists (select 1
from table t1
where t1.name = t.name and t1.mfgdate > t.mfgdate and t1.qtysold < t.qtysold
)
then 1 else 0 end
) over (partition by name) as grp
from table t;
Now you can filter out the names which has increaed sales :
select t.*
from ( <subquery here> ) t
where t.grp = 0;

Answer 4

MySQL Self Join:
SELECT
*
FROM
Bikes AS this_year
LEFT JOIN
Bikes AS prev_year
ON prev_year.name = this_year.name
AND prev_year.mfgdate = DATE_SUB(this_year.mfgdate, INTERVAL 1 YEAR)
GROUP BY
this_year.name
HAVING
MIN(this_year.qtysold - COALESCE(prev_year.qtysold, 0)) > 0
MariaDB LAG():
WITH
historicised AS
(
SELECT
*,
LAG(qtysold) OVER (PARTITION BY name
ORDER BY mfgdate
)
AS prev_qtysold
FROM
Bikes
)
SELECT
name
FROM
historicised
GROUP BY
name
HAVING
MIN(qtysold - COALESCE(prev_qtysold, 0)) > 0