Search code examples
mysqlsqlmaxderived-table

Find which entry occurs maximum number of time in Mysql


I have a table T like this :

ID | brand
2 | 100
2 | 300
2 | 300
1 | 100
1 | 200

I want to find which ID,brand combination occur more than 50% for an ID. For ex : here the result should come as

ID | brand
2 | 300

as (2,300) occurs 2 out of 3 times in ID 2. In ID1, both combinations are 50% times so no row for ID 1.
Now I approached it like this:
First I found count of each ID,brand combination :

(select id,brand,count(*) c from T group by id,brand) as F

Then I list all IDs which satisfy above mentioned criteria

select ID from F group by ID having max(c)/count(*) > 0.5

It works perfectly, but the problem is I can't get corresponding brand because when in above query, I give

select ID, brand ...

Then it always gives first brand of the table.
Here is the sql fiddle


Solution

  • Here's my first attemp.

    SELECT ID, BRAND
    FROM
    (
      SELECT t.ID, t.BRAND, s.TotalCount, 
              COUNT(*) * 1.0 / s.TotalCount Percentage
      FROM   T
             INNER JOIN
            (
              SELECT ID, COUNT(*) totalCOunt
              FROM   T
              GROUP  BY ID
            ) s ON T.ID = s.ID
      GROUP  BY ID, BRAND
      HAVING (COUNT(*) * 1.0 / s.TotalCount) > 0.5
    ) r