feat(db-problems): problems solved, db-backup-tar created #41
Conversation
|
|
||
| -- 2 | ||
|
|
||
| SELECT c.name, |
| INNER JOIN states s ON c.id = s.country_id | ||
| GROUP BY c.name; | ||
| --2 | ||
| SELECT COUNT(*) employees_without_bosses |
src/answers.sql
Outdated
| -- 4 | ||
|
|
||
| SELECT supervisor_id, | ||
| COUNT(supervisor_id) as count |
There was a problem hiding this comment.
Nothing wrong with this one, works as expected. However, I'd have count by employee.id, that would have make the intention a little bit more clear. Anyways, this is totally fine
There was a problem hiding this comment.
Sure, it helps interpretate the query by only reading first SELECT statement!
Done!
src/answers.sql
Outdated
| -- 5 | ||
|
|
||
| SELECT COUNT(*) list_of_office |
There was a problem hiding this comment.
Naming could be improved. I propose offices_in_colorado
| -- 7 | ||
|
|
||
| WITH counts AS ( |
There was a problem hiding this comment.
This query can be significantly simplified. You already have thE CTE, you could have use that table to easily get the one with most and the one with less employees. There is not need of using having.
There was a problem hiding this comment.
I resolved and refractored it, and it makes sense to avoid having statement.
WITH counts AS (
SELECT o.address, COUNT(o.id) AS count
FROM offices o
INNER JOIN employees e ON e.office_id = o.id
GROUP BY o.address
ORDER BY count
)
(
SELECT * FROM counts
ORDER BY count DESC
LIMIT 1
)
UNION ALL
(
SELECT * FROM counts
LIMIT 1
);This solution avoids using having, and only makes 1 inner join. Leading to using less data on the queries after the CTE.
| ) | ||
| LIMIT 1 | ||
| ) | ||
| UNION ALL |
There was a problem hiding this comment.
Why using UNION ALL instead of just UNION?
There was a problem hiding this comment.
For better data normalization.
UNION would gives me only one row in case max and min are the same. However client most like would expect to have the max in the first row and the min on the second row.
It's not that much data where it really matters avoiding duplicates, so having both rows makes sense to me to keep consistency in returning 2 rows each request
src/answers.sql
Outdated
| e.first_name || ' ' || e.last_name as full_name, | ||
| e.email, | ||
| e.job_title, | ||
| o.name company, |
There was a problem hiding this comment.
Try to keep consistency when naming alias. Either use AS always, or don't ever use it
There was a problem hiding this comment.
Done!
SELECT e.uuid,
e.first_name || ' ' || e.last_name as full_name,
e.email,
e.job_title,
o.name as company,
c.name country,
s.name as state,
es.first_name as boss_name
FROM employees e
INNER JOIN offices o ON e.office_id = o.id
INNER JOIN countries c ON c.id = o.country_id
INNER JOIN employees es ON e.supervisor_id = es.id
INNER JOIN states s ON s.id = o.state_id
|
Thanks for addressing the comments. Looks better now |
No description provided.