602. Friend Requests II: Who Has the Most Friends


In social network like Facebook or Twitter, people send friend requests and accept others' requests as well.

Table request_accepted holds the data of friend acceptance, while requester_id and accepter_id both are the id of a person. 

| requester_id | accepter_id | accept_date|
|--------------|-------------|------------|
| 1            | 2           | 2016_06-03 |
| 1            | 3           | 2016-06-08 |
| 2            | 3           | 2016-06-08 |
| 3            | 4           | 2016-06-09 |
Write a query to find the the people who has most friends and the most friends number. For the sample data above, the result is: 
| id | num |
|----|-----|
| 3  | 3   |
Note:
  • It is guaranteed there is only 1 people having the most friends.
  • The friend request could only been accepted once, which mean there is no multiple records with the same requester_id and accepter_id value.

    Explanation:
    The person with id '3' is a friend of people '1', '2' and '4', so he has 3 friends in total, which is the most number than any others.

    Follow-up:
    In the real world, multiple people could have the same most number of friends, can you find all these people in this case?

    • b'
    \n
    \n\n
    \n

    Solution

    \n
    \n

    Approach: Union requester_id and accepter_id [Accepted]

    \n

    Algorithm

    \n

    Being friends is bidirectional, so if one person accepts a request from another person, both of them will have one more friend.

    \n

    Thus, we can union column requester_id and accepter_id, and then count the number of the occurrence of each person.

    \n
    select requester_id as ids from request_accepted\nunion all\nselect accepter_id from request_accepted;\n
    \n
    \n

    Note: Here we should use union all instead of union because union all will keep all the records even the \'duplicated\' one.

    \n
    \n

    Taking the sample as an example, the output is:

    \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
    ids
    1
    1
    2
    3
    2
    3
    3
    4
    \n

    Then it will be fairly easy to get the \'ids\' with most occurrence using the same technique as mentioned in problem 580. Customer Placing the Largest Number of Orders.

    \n

    MySQL

    \n
    select ids as id, cnt as num\nfrom\n(\nselect ids, count(*) as cnt\n   from\n   (\n        select requester_id as ids from request_accepted\n        union all\n        select accepter_id from request_accepted\n    ) as tbl1\n   group by ids\n   ) as tbl2\norder by cnt desc\nlimit 1\n;\n
    \n
    '