I have two tables (country & ducks) where the country table has every country in the world and the ducks table has a list of ducks with a country_id field to link to the
Many databases have some equivalent of "select top 10 * from...". In mySql, the syntax would be "select * from ... limit 10".
... BUT ...
In this case, uou really want "group by" and "max()"!
You might try:
SELECT c.*, d.*
FROM country c
INNER JOIN (
SELECT d.country_id, d.id, MAX(d.rating) AS rating
FROM ducks d
GROUP BY d.country_id
) q ON (q.country_id = c.id)
INNER JOIN ducks d
ON (d.country_id, d.rating) = (q.country_id, q.rating)
SELECT c.*, d.*
FROM country c
INNER JOIN ducks d
ON d.id = --- guessing the ducks Primary Key here
( SELECT dd.id --- and here
FROM ducks dd
WHERE c.id = dd.country_id
ORDER BY dd.rating DESC
LIMIT 1
)
An index on (country_id, rating, id)
for MyISAM table or (country_id, rating)
for InnoDB table, would help.
This query will show only one duck
per country, even with more than one having the same rating. If you want ducks with tied rating to appear, use @imm's GROUP BY
answer.
Try this:
SELECT c.country, MAX(d.rating) AS max_rating
FROM country c
JOIN ducks d ON c.id = d.country_id
GROUP BY c.id
ORDER BY c.country ASC
If the "highest rating" is 1, then change MAX(d.rating)
to MIN(d.rating)
You could try just adding a selecting join, for
SELECT c.*, d.*
FROM country c
INNER JOIN ducks d ON c.id = d.country_id
LEFT JOIN ducks d2 ON d.country_id = d2.country_id AND d2.rating > d.rating
WHERE d2.id IS NULL