Table: Traffic
+---------------+---------+
| Column Name | Type |
+---------------+---------+
| user_id | int |
| activity | enum |
| activity_date | date |
+---------------+---------+
There is no primary key for this table, it may have duplicate rows.
The activity column is an ENUM type of ('login', 'logout', 'jobs', 'groups', 'homepage').
Write an SQL query that reports for every date within at most 90 days from today, the number of users that logged in for the first time on that date. Assume today is 2019-06-30.
The query result format is in the following example:
Traffic table: +---------+----------+---------------+ | user_id | activity | activity_date | +---------+----------+---------------+ | 1 | login | 2019-05-01 | | 1 | homepage | 2019-05-01 | | 1 | logout | 2019-05-01 | | 2 | login | 2019-06-21 | | 2 | logout | 2019-06-21 | | 3 | login | 2019-01-01 | | 3 | jobs | 2019-01-01 | | 3 | logout | 2019-01-01 | | 4 | login | 2019-06-21 | | 4 | groups | 2019-06-21 | | 4 | logout | 2019-06-21 | | 5 | login | 2019-03-01 | | 5 | logout | 2019-03-01 | | 5 | login | 2019-06-21 | | 5 | logout | 2019-06-21 | +---------+----------+---------------+ Result table: +------------+-------------+ | login_date | user_count | +------------+-------------+ | 2019-05-01 | 1 | | 2019-06-21 | 2 | +------------+-------------+ Note that we only care about dates with non zero user count. The user with id 5 first logged in on 2019-03-01 so he's not counted on 2019-06-21.
My solution
Db-fiddle for this problem
My idea was to rank the login date for each user and select the very 1st login date (rank=1) for each user in the subquery.
select temp.activity_date as login_date,count(temp.user_id) as user_count
from
(select user_id,activity_date,rank() over (partition by user_id order by activity_date) as ranking
from Traffic
where activity='login'
group by user_id,activity_date) temp
where temp.ranking=1
group by temp.activity_date
having datediff('2019-06-30',temp.activity_date) < 90