Compter les lignes avec des tables jointes internes
J'ai 3 tables:
Joueurs:
mysql> SELECT * FROM players;
+-----------+---------+----------------------+----------------------+-----------------+------------------------------+-----------------------+---------------------+
| player_id | team_id | player_name | player_jersey_number | player_position | player_email | player_contact_number | player_timestamp |
+-----------+---------+----------------------+----------------------+-----------------+------------------------------+-----------------------+---------------------+
| 1 | 4 | Popoy Alfonso | 2 | | [email protected] | 09263453234 | 2015-08-05 00:48:10 |
| 2 | 4 | Karlo Ripas | 10 | | [email protected] | 09212354324 | 2015-08-05 00:50:03 |
| 3 | 4 | VHaughn Von | 32 | | [email protected] | 09361234565 | 2015-08-05 00:51:00 |
| 4 | 4 | Lordie Zalbahe | 23 | | [email protected] | 09391222334 | 2015-08-05 00:52:42 |
| 5 | 4 | Jigs Selda | 8 | | [email protected] | 09325566323 | 2015-08-05 00:53:36 |
| 6 | 4 | Rhan Garniel | 3 | | [email protected] | 09129503400 | 2015-08-05 00:54:20 |
| 7 | 5 | Johnritz Rodriguez | 11 | | [email protected] | 09231112346 | 2015-08-05 00:56:02 |
| 8 | 5 | Garret Van Sarmiento | 7 | | [email protected] | 09264565600 | 2015-08-05 00:56:53 |
| 9 | 5 | Lester Selda Lineses | 12 | | [email protected] | 09068746354 | 2015-08-05 00:57:47 |
| 10 | 5 | Laurence Lineses | 44 | | [email protected] | 09847354672 | 2015-08-05 00:59:33 |
| 11 | 5 | Xandrix Buendia | 1 | | [email protected] | 09234665590 | 2015-08-05 01:00:12 |
| 12 | 5 | Betoyskie Limpiada | 45 | | [email protected] | 09213456667 | 2015-08-05 01:01:15 |
+-----------+---------+----------------------+----------------------+-----------------+------------------------------+-----------------------+---------------------+
Équipes:
mysql> SELECT * FROM teams;
+---------+-----------+----------------------+---------------------+
| team_id | season_id | team_name | team_timestamp |
+---------+-----------+----------------------+---------------------+
| 4 | 1 | Quiapo A | 2015-08-05 00:30:13 |
| 5 | 1 | Quiapo B | 2015-08-05 00:30:25 |
| 6 | 1 | Balik-Balik Warriors | 2015-08-05 00:31:13 |
| 7 | 1 | Adamson Falcons | 2015-08-05 00:31:42 |
| 8 | 1 | Pasay Flooders | 2015-08-05 00:32:04 |
| 9 | 1 | Marina Dragons | 2015-08-05 00:32:22 |
| 10 | 1 | MDC Archers | 2015-08-05 00:33:12 |
| 11 | 2 | Quiapo A | 2015-08-05 00:34:25 |
| 12 | 2 | Quiapo B | 2015-08-05 00:34:38 |
| 13 | 2 | Marikina Eagels | 2015-08-05 00:35:11 |
| 14 | 2 | TIP Steallers | 2015-08-05 00:35:32 |
| 15 | 2 | Gasan Blue Eagles | 2015-08-05 00:36:12 |
+---------+-----------+----------------------+---------------------+
Saisons:
mysql> SELECT * FROM seasons;
+-----------+-------------+----------------------+---------------------+
| season_id | season_name | season_event_name | season_timestamp |
+-----------+-------------+----------------------+---------------------+
| 1 | Season 1 | Summer Games | 2015-08-05 00:23:15 |
| 2 | Season 2 | Aniversary Sportfest | 2015-08-05 00:25:10 |
+-----------+-------------+----------------------+---------------------+
Je travaille actuellement sur cette requête mais le résultat n'est pas correct.
SELECT
teams.team_name,
(
SELECT COUNT(*)
FROM teams
INNER JOIN players
ON teams.team_id = players.team_id
) as num_of_players,
teams.team_timestamp
FROM teams
INNER JOIN seasons
ON seasons.season_id = teams.season_id
GROUP BY teams.team_name;
Production:
+----------------------+----------------+---------------------+
| team_name | num_of_players | team_timestamp |
+----------------------+----------------+---------------------+
| Adamson Falcons | 12 | 2015-08-05 00:31:42 |
| Balik-Balik Warriors | 12 | 2015-08-05 00:31:13 |
| Gasan Blue Eagles | 12 | 2015-08-05 00:36:12 |
| Marikina Eagels | 12 | 2015-08-05 00:35:11 |
| Marina Dragons | 12 | 2015-08-05 00:32:22 |
| MDC Archers | 12 | 2015-08-05 00:33:12 |
| Pasay Flooders | 12 | 2015-08-05 00:32:04 |
| Quiapo A | 12 | 2015-08-05 00:30:13 |
| Quiapo B | 12 | 2015-08-05 00:30:25 |
| TIP Steallers | 12 | 2015-08-05 00:35:32 |
+----------------------+----------------+---------------------+
Le résultat que je veux est le suivant:
+----------------------+----------------+---------------------+
| team_name | num_of_players | team_timestamp |
+----------------------+----------------+---------------------+
| Adamson Falcons | 0 | 2015-08-05 00:31:42 |
| Balik-Balik Warriors | 0 | 2015-08-05 00:31:13 |
| Gasan Blue Eagles | 0 | 2015-08-05 00:36:12 |
| Marikina Eagels | 0 | 2015-08-05 00:35:11 |
| Marina Dragons | 0 | 2015-08-05 00:32:22 |
| MDC Archers | 0 | 2015-08-05 00:33:12 |
| Pasay Flooders | 0 | 2015-08-05 00:32:04 |
| Quiapo A | 6 | 2015-08-05 00:30:13 |
| Quiapo B | 6 | 2015-08-05 00:30:25 |
| TIP Steallers | 0 | 2015-08-05 00:35:32 |
+----------------------+----------------+---------------------+
Réponse complètement réarrangée.
J'ai ajouté des données aux tables des joueurs et des équipes pour rendre les réponses plus générales - voir le bas du message pour tous les DDL (CREATE TABLE tab_name...) et DML (INSERT INTO tab_name VALUES...) utilisé dans cette réponse. J'ai également créé la table des saisons (inchangée par rapport aux données originales de l'OP - c'est-à-dire vos).
BTW, bienvenue sur le forum. Mais vous auriez vraiment dû nous donner le DDL et le DML. Jetez un coup d'œil à la visite et au blog "comment nous aider à vous aider" - tous deux en bas à gauche de la page. Mais je me suis intéressé et je l'ai fait moi-même, mais vous obtiendrez plus de personnes pour vous aider si vous fournissez DDL et DML.
Après avoir créé et chargé les tables, j'ai exécuté le SQL suivant.
SELECT t1.team_name,
IFNULL(t2.num_players, 0) AS strength,
t1.team_timestamp
FROM team t1
LEFT OUTER JOIN
(SELECT team_id, COUNT(team_id) AS num_players
FROM player
GROUP BY team_id
) t2
ON t1.team_id = t2.team_id
-- GROUP BY t1.team_id, t1.team_name, t1.season_id -- **NOTE** - see discussion below
ORDER BY strength DESC, team_name ASC;
Et le résultat n'est pas tout à fait celui que vous voulez, mais proche.
+----------------------+----------+---------------------+
| team_name | strength | team_timestamp |
+----------------------+----------+---------------------+
| Quiapo B | 7 | 2015-08-05 00:30:25 |
| Quiapo B | 7 | 2015-08-05 00:30:25 |
| Quiapo A | 6 | 2015-08-05 00:30:13 |
| Quiapo A | 6 | 2015-08-05 00:30:13 |
| Adamson Falcons | 0 | 2015-08-05 00:31:42 |
| Balik-Balik Warriors | 0 | 2015-08-05 00:31:13 |
| Gasan Blue Eagles | 0 | 2015-08-05 00:36:12 |
| Marikina Eagels | 0 | 2015-08-05 00:35:11 |
| Marina Dragons | 0 | 2015-08-05 00:32:22 |
| MDC Archers | 0 | 2015-08-05 00:33:12 |
| Pasay Flooders | 0 | 2015-08-05 00:32:04 |
| TIP Steallers | 0 | 2015-08-05 00:35:32 |
+----------------------+----------+---------------------+
L'exécution de la requête oNare donne (!)
+-----------+----------------+---------------------+
| team_name | num_of_players | team_timestamp |
+-----------+----------------+---------------------+
| Quiapo A | 26 | 2015-08-05 00:30:13 |
+-----------+----------------+---------------------+
Notez que 26 = 2 * (6 + 7). Mais si sql_mode est réglé sur STRICT_TRANS_TABLES,NO_ENGINE_SUBSTITUTION,ONLY_FULL_GROUP_BY, cette requête échouera avec le message
ERROR 1140 (42000): Mixing of GROUP columns (MIN(),MAX(),COUNT(),...) with no GROUP columns is illegal if there is no GROUP BY clause
Cependant, une légère modification de la requête oNare donne les mêmes résultats que ceux obtenus avec ma première requête (avec sql_mode non défini sur ONLY_FULL_GROUP_BY).
SELECT
team.team_name,
COUNT(player.player_id) as num_of_players,
team.team_timestamp
FROM team
LEFT JOIN player ON (player.team_id = team.team_id)
LEFT JOIN seasons ON (seasons.season_id = team.season_id)
GROUP BY team.team_name, team.season_id, team.team_timestamp -- **ADDED**
ORDER BY num_of_players DESC, team.team_name ASC;
L'ajout de GROUP BY donne la bonne réponse. Il est intéressant de noter que PostgreSQL générera une erreur si le GROUP BY la ligne est mise en commentaire. Ma requête ne fonctionnera pas non plus sur PostgreSQL si je n'ajoute pas de GROUP BY ligne juste avant le ORDER BY ligne.
Je pense que comme réponse générale, vous pouvez accepter l'une de nos réponses comme correcte pour votre question (avec GROUP BY) - il n'est pas possible d'obtenir les résultats souhaités sans plus d'informations dans les tableaux, mais voir ci-dessous.
En aparté, ONLY_FULL_GROUP_BY sera le mode par défaut pour 5.7 , aussi bien l'obtenir dès maintenant!
Je pense que la structure de votre table a besoin d'une révision. Vous devriez avoir des joueurs pour différentes saisons (cela a du sens, le transfert de personnes). Vous devriez avoir les mêmes ids pour la même équipe sur différentes saisons - sinon comment allez-vous agréger les statistiques (buts/points/matchs gagnés/perdus) sur plusieurs saisons?
TABLEAU DDL et DML - STRUCTURE et CONTENU.
CREATE TABLE player (player_id INT, team_id INT, player_name VARCHAR(25));
CREATE TABLE team( team_id INT, season_id INT, team_name VARCHAR(25), team_timestamp TIMESTAMP);
CREATE TABLE season (season_id INT, season_name VARCHAR(25));
INSERT INTO player VALUES (1, 4, 'Popoy Alfonso');
INSERT INTO player VALUES (2, 4, 'Karlo Ripas');
INSERT INTO player VALUES (3, 4, 'VHaughn Von');
INSERT INTO player VALUES (4, 4, 'Lordie Zalbahe');
INSERT INTO player VALUES (5, 4, 'Jigs Selda' );
INSERT INTO player VALUES (6, 4, 'Rhan Garniel' );
INSERT INTO player VALUES (7, 5, 'Johnritz Rodriguez' );
INSERT INTO player VALUES (8, 5, 'Garret Van Sarmiento');
INSERT INTO player VALUES (9, 5, 'Lester Selda Lineses');
INSERT INTO player VALUES (10, 5, 'Laurence Lineses' );
INSERT INTO player VALUES (11, 5, 'Xandrix Buendia' );
INSERT INTO player VALUES (12, 5, 'Betoyskie Limpiada' );
INSERT INTO player VALUES (12, 5, 'Donald Duck' ); -- Added Donald Duck!
-- I inserted extra records into the player table as below.
INSERT INTO player VALUES (1, 11, 'Popoy Alfonso');
INSERT INTO player VALUES (2, 11, 'Karlo Ripas');
INSERT INTO player VALUES (3, 11, 'VHaughn Von');
INSERT INTO player VALUES (11, 11, 'Lordie Zalbahe');
INSERT INTO player VALUES (5, 11, 'Jigs Selda' );
INSERT INTO player VALUES (6, 11, 'Rhan Garniel' );
INSERT INTO player VALUES (7, 12, 'Johnritz Rodriguez' );
INSERT INTO player VALUES (8, 12, 'Garret Van Sarmiento');
INSERT INTO player VALUES (9, 12, 'Lester Selda Lineses');
INSERT INTO player VALUES (10, 12, 'Laurence Lineses' );
INSERT INTO player VALUES (11, 12, 'Xandrix Buendia' );
INSERT INTO player VALUES (12, 12, 'Betoyskie Limpiada' );
Pour la table des équipes, j'ai donné à QuiradoA et QuiradoB les mêmes id dans la saison 1 et la saison 2 - cela n'a aucun sens pour moi d'avoir des ids différents pour la même équipe juste parce que la saison change. J'ai également conservé le même TIMESTAMP pour QuiradoA et B.
INSERT INTO team VALUES ( 4 , 1 , 'Quiapo A' , '2015-08-05 00:30:13');
INSERT INTO team VALUES ( 5 , 1 , 'Quiapo B' , '2015-08-05 00:30:25');
INSERT INTO team VALUES ( 6 , 1 , 'Balik-Balik Warriors' , '2015-08-05 00:31:13');
INSERT INTO team VALUES ( 7 , 1 , 'Adamson Falcons' , '2015-08-05 00:31:42');
INSERT INTO team VALUES ( 8 , 1 , 'Pasay Flooders' , '2015-08-05 00:32:04');
INSERT INTO team VALUES ( 9 , 1 , 'Marina Dragons' , '2015-08-05 00:32:22');
INSERT INTO team VALUES ( 10 , 1 , 'MDC Archers' , '2015-08-05 00:33:12');
INSERT INTO team VALUES ( 4 , 2 , 'Quiapo A' , '2015-08-05 00:30:13'); -- **NOTE** `id` and `TIMESTAMP` for both A & B.
INSERT INTO team VALUES ( 5 , 2 , 'Quiapo B' , '2015-08-05 00:30:25');
INSERT INTO team VALUES ( 13 , 2 , 'Marikina Eagels' , '2015-08-05 00:35:11');
INSERT INTO team VALUES ( 14 , 2 , 'TIP Steallers' , '2015-08-05 00:35:32');
INSERT INTO team VALUES ( 15 , 2 , 'Gasan Blue Eagles' , '2015-08-05 00:36:12');
INSERT INTO season VALUES(1, 'Season 1');
INSERT INTO season VALUES(2, 'Season 2');
Votre requête vous donne 12 num_of_players Parce que votre comptage ne fait que renvoyer les sous-requêtes, si vous exécutez SELECT COUNT(*) FROM teams INNER JOIN players ON teams.team_id = players.team_id; vous verrez ce que vous faites vraiment.
Pour corriger votre syntaxe juste un de plus LEFT JOIN:
SELECT
teams.team_name,
COUNT(players.player_id) as num_of_players,
teams.team_timestamp
FROM test.teams
LEFT JOIN test.players ON (players.team_id=teams.team_id)
LEFT JOIN test.seasons ON (seasons.season_id = teams.season_id)
GROUP BY teams.team_name;
Informations sur le test:
mysql> SELECT * FROM test.players;
+-----------+---------+--------------------+----------------------+-----------------+--------------+-----------------------+------------------+
| player_id | team_id | player_name | player_jersey_number | player_position | player_email | player_contact_number | player_timestamp |
+-----------+---------+--------------------+----------------------+-----------------+--------------+-----------------------+------------------+
| 1 | 4 | Popoy Alfonso | 2 | NULL | NULL | NULL | NULL |
| 2 | 4 | NULL | 10 | NULL | NULL | NULL | NULL |
| 3 | 4 | NULL | NULL | NULL | NULL | NULL | NULL |
| 4 | 4 | NULL | NULL | NULL | NULL | NULL | NULL |
| 5 | 4 | NULL | NULL | NULL | NULL | NULL | NULL |
| 6 | 4 | NULL | NULL | NULL | NULL | NULL | NULL |
| 7 | 5 | Johnritz Rodriguez | 11 | NULL | NULL | NULL | NULL |
| 8 | 5 | NULL | NULL | NULL | NULL | NULL | NULL |
| 9 | 5 | NULL | NULL | NULL | NULL | NULL | NULL |
| 10 | 5 | NULL | NULL | NULL | NULL | NULL | NULL |
| 11 | 5 | NULL | NULL | NULL | NULL | NULL | NULL |
| 12 | 5 | NULL | NULL | NULL | NULL | NULL | NULL |
+-----------+---------+--------------------+----------------------+-----------------+--------------+-----------------------+------------------+
12 rows in set (0.00 sec)
mysql> SELECT * FROM test.teams;
+---------+-----------+----------------------+----------------+
| team_id | season_id | team_name | team_timestamp |
+---------+-----------+----------------------+----------------+
| 4 | 1 | Quiapo A | NULL |
| 5 | 1 | Quiapo B | NULL |
| 6 | 1 | Balik-Balik Warriors | NULL |
| 7 | 1 | Adamson Falcons | NULL |
| 8 | 1 | Pasay Flooders | NULL |
| 9 | 1 | Marina Dragons | NULL |
| 10 | 1 | MDC Archers | NULL |
| 11 | 2 | Quiapo A | NULL |
| 12 | 2 | Quiapo B | NULL |
| 13 | 2 | Marikina Eagels | NULL |
| 14 | 2 | TIP Steallers | NULL |
| 15 | 2 | Gasan Blue Eagles | NULL |
+---------+-----------+----------------------+----------------+
12 rows in set (0.00 sec)
mysql> SELECT * FROM test.seasons;
+-----------+-------------+----------------------+------------------+
| season_id | season_name | season_event_name | season_timestamp |
+-----------+-------------+----------------------+------------------+
| 1 | Season 1 | Summer Games | NULL |
| 2 | Season 2 | Aniversary Sportfest | NULL |
+-----------+-------------+----------------------+------------------+
2 rows in set (0.00 sec)
mysql>
Votre REQUÊTE:
mysql> SELECT teams.team_name, (SELECT COUNT(*) FROM teams INNER JOIN players ON teams.team_id = players.team_id) as num_of_players, teams.team_timestamp
-> FROM teams
-> INNER JOIN seasons
-> ON seasons.season_id = teams.season_id
-> GROUP BY teams.team_name;
+----------------------+----------------+----------------+
| team_name | num_of_players | team_timestamp |
+----------------------+----------------+----------------+
| Adamson Falcons | 12 | NULL |
| Balik-Balik Warriors | 12 | NULL |
| Gasan Blue Eagles | 12 | NULL |
| Marikina Eagels | 12 | NULL |
| Marina Dragons | 12 | NULL |
| MDC Archers | 12 | NULL |
| Pasay Flooders | 12 | NULL |
| Quiapo A | 12 | NULL |
| Quiapo B | 12 | NULL |
| TIP Steallers | 12 | NULL |
+----------------------+----------------+----------------+
10 rows in set (0.00 sec)
mysql>
Requête fixe:
mysql> SELECT
-> teams.team_name,
-> COUNT(players.player_id) as num_of_players,
-> teams.team_timestamp
-> FROM test.teams
-> LEFT JOIN test.players ON (players.team_id=teams.team_id)
-> LEFT JOIN test.seasons ON (seasons.season_id = teams.season_id)
-> GROUP BY teams.team_name;
+----------------------+----------------+----------------+
| team_name | num_of_players | team_timestamp |
+----------------------+----------------+----------------+
| Adamson Falcons | 0 | NULL |
| Balik-Balik Warriors | 0 | NULL |
| Gasan Blue Eagles | 0 | NULL |
| Marikina Eagels | 0 | NULL |
| Marina Dragons | 0 | NULL |
| MDC Archers | 0 | NULL |
| Pasay Flooders | 0 | NULL |
| Quiapo A | 6 | NULL |
| Quiapo B | 6 | NULL |
| TIP Steallers | 0 | NULL |
+----------------------+----------------+----------------+
10 rows in set (0.00 sec)
mysql>
Voir la sortie ici: SQLFiddle