MYSQL-GROUP_CONCAT仅显示重复项

Question

I have 2 SQL tables.

1. The first table has a mapping of department-ids to student-ids. (A student may belong to more than one department and a department has many students).
2. The second table has a mapping of students and their hobbies. (which also has a many to many relationship)

Sample Snapshots of tables:

DESC DEPARTMENT_STUDENTS;
----------------------------
    DEPT_ID    STUDENT_ID
----------------------------
    Physics       123
    Mathematics   111
    Physics       111
    CS            45
    CS            56
    Mathematics   89

DESC STUDENTS_HOBBIES;
-------------------------------
    STUDENT_ID      HOBBY
-------------------------------
       111          Skiing
       111          Singing
       111          Browsing
       123          Singing
       123          Browsing
       123          Reading
       45           Origami
       56           Origami
       56           Making Prank Calls
       89           Reading

I am looking for a query that will provide a list of common hobbies among all the students per department. Required output would look like :

-----------------------------------------
     DEPT_ID         GROUP_CONCAT(...)
-----------------------------------------
    Physics          (Singing, Browsing)
    Mathematics      ()
    CS               (Origami)

I am almost there and after a little bit of fiddling around with MySQL, I was able to group ALL(not the common ones) the hobbies of a department. GROUP_CONCAT provides ways to eliminate duplicates also. One way that I could think of , is to retrieve duplicates programatically over the query below:

SELECT DEPT_ID, GROUP_CONCAT(OP.HOBBIES) FROM DEPARTMENT_STUDENTS DS
   INNER JOIN (SELECT STUDENT_ID, GROUP_CONCAT(HOBBY)AS HOBBIES FROM STUDENTS_HOBBIES 
               GROUP  BY STUDENT_ID) OP
           ON OP.STUDENT_ID = DS.STUDENT_ID
GROUP  BY DS.DEPT_ID; 

Is there a way to preserve duplicates alone? Even in this query, any optimizations are welcome. Thanks!

Answer 1

Here you go:

SELECT DEPT_ID, GROUP_CONCAT(HOBBY) FROM (
SELECT 
d.DEPT_ID
, s.HOBBY
FROM
department_students d
INNER JOIN students_hobbies s USING(STUDENT_ID)
GROUP BY d.DEPT_ID, s.HOBBY
HAVING COUNT(DISTINCT s.STUDENT_ID) > 1
)sq 
GROUP BY DEPT_ID