How to write the subquery? - database

Question

I am no sure how to do the following query for the following tables:
1st table - book_type:
book_type_id
book_type_code
book_type_description
2nd table - books:
book_id
book_type_id
book_code
book_status
book_description
3rd table - sold_books:
book_id
price
sold_books status (canceled, or not)
amount_of_books_sold
Books that have an average of amount_of_books_sold that is less than the average of amount_of_books_sold for the book_type that the book is in.
Do not show canceled books from sold_books_status.
Only these columns:
book_type_description, book_type_code, book_description
avg amount_of_books_sold, avg amount_of_books_sold forsook_type
First I tried to do all in one query. Then I tried using with statement. It didn't show me an error, but nothing was shown.
Then I tried with subqueries. No success, because I didn't know what to put in subquery.
Any ideas how to create this query or subquery in sqlite ?

Answer 1

There're a few things need to be clarified:
What's the relationship between books and sold_books?
Does avg_amount_of_books_sold include sold_book_status='canceled'?
Assume that
books and sold_books is 1-to-1 (This means book_id in sold_books is a primary key and also a foreign key).
avg_amount_of_books_sold does not include sold_book_status='canceled'
Then we can do this in a few steps:
Step 1. Calculate avg_amount_of_books_sold (please remove where... clause if canceled books are included.)
select avg(amount_of_books_sold) as avg_amount_of_books_sold
from sold_books
where sold_book_status <> 'canceled';
Step 2. Join all three tables to get those columns you need.
select t.book_type_description,
t.book_type_code,
b.book_description,
s.amount_of_books_sold
from book_type t
join books b
on t.book_type_id = b.book_type_id
join sold_books s
on b.book_id = s.book_id
where s.sold_book_status <> 'canceled';
Step 3. Put the above two queries in CTE sub-query and then compares amount_of_books_sold and avg_amount_of_books_sold in where clause to include only amount_of_books_sold < avg_amount_of_books_sold.
with cte_avg_amount_of_books_sold as (
select avg(amount_of_books_sold) as avg_amount_of_books_sold
from sold_books
where sold_book_status <> 'canceled'),
cte_sold_books_not_canceled as (
select t.book_type_description,
t.book_type_code,
b.book_description,
s.amount_of_books_sold
from book_type t
join books b
on t.book_type_id = b.book_type_id
join sold_books s
on b.book_id = s.book_id
where s.sold_book_status <> 'canceled')
select t.*, a.*
from cte_sold_books_not_canceled t, cte_avg_amount_of_books_sold a
where t.amount_of_books_sold < a.avg_amount_of_books_sold;
Please note that if the relationship between books and sold_books is not 1-to-1 but 1-to-many, then you need to have a rule about either how to pick a row from sold_books for a specific book_id or how to aggregate the information in sold_books to book_id level.