mysql - How to delete duplicates in SQL table based on multiple fields

Question

I have a table of games, which is described as follows:

+---------------+-------------+------+-----+---------+----------------+
| Field         | Type        | Null | Key | Default | Extra          |
+---------------+-------------+------+-----+---------+----------------+
| id            | int(11)     | NO   | PRI | NULL    | auto_increment |
| date          | date        | NO   |     | NULL    |                |
| time          | time        | NO   |     | NULL    |                |
| hometeam_id   | int(11)     | NO   | MUL | NULL    |                |
| awayteam_id   | int(11)     | NO   | MUL | NULL    |                |
| locationcity  | varchar(30) | NO   |     | NULL    |                |
| locationstate | varchar(20) | NO   |     | NULL    |                |
+---------------+-------------+------+-----+---------+----------------+

But each game has a duplicate entry in the table somewhere, because each game was in the schedules for two teams. Is there a sql statement I can use to look through and delete all the duplicates based on identical date, time, hometeam_id, awayteam_id, locationcity, and locationstate fields?

Answer 1

You should be able to do a correlated subquery to delete the data. Find all rows that are duplicates and delete all but the one with the smallest id. For MYSQL, an inner join (functional equivalent of EXISTS) needs to be used, like so:

delete games from games inner join 
    (select  min(id) minid, date, time,
             hometeam_id, awayteam_id, locationcity, locationstate
     from games 
     group by date, time, hometeam_id, 
              awayteam_id, locationcity, locationstate
     having count(1) > 1) as duplicates
   on (duplicates.date = games.date
   and duplicates.time = games.time
   and duplicates.hometeam_id = games.hometeam_id
   and duplicates.awayteam_id = games.awayteam_id
   and duplicates.locationcity = games.locationcity
   and duplicates.locationstate = games.locationstate
   and duplicates.minid <> games.id)

To test, replace delete games from games with select * from games. Don't just run a delete on your DB :-)