Question

I have a table called releases, part of a ‘music’ database. The table consists of two columns; cd_id, band_id

I want to list the cd_id and the number of bands involved for each cd where two or more band_id are associated with one cd_id value.

Could you help me phrase this? Thank you in advance.

Answer 1

This is an aggregate COUNT() with a HAVING clause:

SELECT
  cd_id,
  count(*) AS numbands
FROM releases
GROUP BY cd_id
HAVING numbands >= 2

The COUNT(*) coupled with the GROUP BY cd_id returns the number of band_id per cd_id, and in order to limit these to releases with 2+ band_id, a HAVING clause specifies numbands (the alias given to the count value) >= 2.

Note, this assumes each band_id is only listed once per cd_id. If that isn’t the case, to get unique bands per cd, use

COUNT(DISTINCT band_id) AS numbands

instead of COUNT(*) AS numbands.