如何查询与其他列的聚合数据

Question

Given is below record in user_score table where each user can have multiple score, user will diff user id can have same name and role. How to fetch average score along with user_id , role and name . I want to group by user_id .

So for my below table I am expecting result as:

Name   Role  Score  user_id
Jack   CEO     6      123
Jack   CEO     2      456
Tony   Dev     5      234
Sham    QA     3      678

user_score table data :

id     name    role     score    user_id
1      Jack    CEO       4        123
2      Jack    CEO       8        123
3      Tony    Dev       9        234
4      Jack    CEO       2        456
5      Tony    Dev       1        234
6      Sham    QA        3        678

Answer 1

Simply use group by :

select id, name, role, avg(score), user_id
from user_score
group by id, name, role, user_id

PS: You can normalize the data because 3/5 columns are redundant (they're unnecessarily repeated). I would leave only score and user_id in this table, move user_id , role and name to the other and use user_id as a joining key.

Answer 2

While you might know that a given user_id always has the same name and role, PostgreSQL has no way of knowing that (other than you normalizing your data), so it needs instructions on what to report for more than one name within a given user_id. You need to include a dummy aggregation on the extra columns.

select min(name) as name, min(role) as role, avg(score) as score, user_id 
    from user_score group by user_id

Rather than min , you could use max or string_agg(distinct ...) , or probably several other things