Retention of users on an Ott platform

user_13949360 0
Viewed 84

I have table called user_info

It has two columns:

User_id
Date

How do I get to the table like the one mentioned below:

----------------------------------------------------------------
Date        |total_number_of_users | D2    |D5  |  D7 |   D14|
--------------------------------------------------------------
2020-07-01      1000                 700    500   200    150
2020-07-02      400                  300    250   200    100

For example consider the first row in the table I am trying to achieve:

total_number_of_users = Total number of users who have visited the site on 2020-07- 01

D2=Out of total users who visited the site on 2020-07-01, visited on 2020-07-2

D7=Out of total users who visited the site on 2020-07-01, visited on 2020-07-7

I have tried the following, how do I get the exact the solution:


SELECT user_id, week(login_date) AS login_week
    FROM user_info
    GROUP BY user_id,week(login_date);
SELECT user_id, min(week(login_date)) AS first_week
       FROM user_info
       GROUP BY user_id;
select a.user_id,a.login_week,b.first_week as first_week  from   
              (SELECT
                user_id,
                week(login_date) AS login_week
                FROM user_info
                GROUP BY user_id,week(login_date)) a,
              (SELECT
                user_id,
                min(week(login_date)) AS first_week
                FROM user_info
                GROUP BY user_id) b
        where a.user_id=b.user_id;
data-science mysql sql
1 Answers

This seems painful, but you can use a self-join and aggregation:

select t.date,
       sum( t2.date = t.date) as total_number_of_users,
       sum( t2.date = t.date + interval 1 day ) as d2,
       sum( t2.date = t.date + interval 4 day ) as d5,
       sum( t2.date = t.date + interval 6 day ) as d7,
       sum( t2.date = t.date + interval 13 day ) as d14
from (select distinct date, user_id
      from t
     ) t1 left join
     (select distinct date, user_id
      from t
     ) t2
     on t1.user_id = t2.user_id and
        t2.date in (t1.date, t1.date + interval 1 day, t1.date + interval 4 day, t1.day + interval 6 day, t1.day + interval 13 day)
group by t.date;
user_11440350
Related