Score interview (1)----related to business

In the interview process of data analysis, the question of user retention rate is often asked. This section will focus on explaining this problem. First, we must understand the definition of retention and know how to calculate retention.Let's start with a small case from the Internet.

Given the user_behavior table, ask to query the user retention rate for the next day, 7th and 30th.

First of all, what is the definition of retention?

It is now set that the first login time of a new user is the 0th day, the new user is defined as the user who logs in for the first time, and the code name of the login behavior is 1.

Next day retention rate: Among the new users on day 0, the number of users who still logged in on the first day after the new day / the total number of new users on day 0;

7-day retention rate: Day 0 new users, the number of users who are still logged in on the 7th day after the new day/ Day 0Add the total number of users;

30-day retention rate: In the new users on the 0th day, the number of users who are still logged in on the 30th day after the new day/Day 0>The total number of new users;

Note: Retention is generally a discrete concept and does not require users to log in every day for N days    

• Select daily new users
• Lists the date each new user logged in for the first time and the dates they have logged in after that date
• Calculate the difference between the listed login dates. If there is a difference of 1 day, it means that the new user is still available for the next day. If there is a difference of 7 days, it means that the new user is still available for seven days, and so on
• Count the number of retained new users every day and calculate the retention rate

  `user_behavior`