如何根据 sqlite 中的条件计算不同的值？

Question

我的名为 matchEvent 的表如下所示：

 eventId    playerId  matchId   eventName description
 --------------------------------------------------------
 100        50        1         goal      freeKick
 100        50        1         goal      shot
 200        50        2         foul      penalty
 300        51        3         goal      header
 300        51        3         goal      ownGoal
 400        51        4         goal      ownGoal
 400        51        4         goal      shot

我想计算特定球员在特定比赛中的所有常规进球（但不包括乌龙球）。 我试过了：

con = sqlite3.connect(my_db)
regular_goals = con.execute("SELECT playerId, matchId, COUNT(DISTINCT eventId) FROM matchEvent WHERE eventName = 'goal' GROUP BY playerId, matchId HAVING description <> 'ownGoal'").fetchall()

问题是 HAVING 子句似乎是根据组的顶行值过滤组，所以它给了我：

[(50, 1, 1),(51, 3, 1)]

而我期望的结果是：

[(50, 1, 1)] 

因为我的表只包含一个常规目标。

Answer 1

您需要 2 个级别的聚合：

SELECT playerId, matchId, SUM(counter) total_goals
FROM (
  SELECT playerId, matchId, COUNT(DISTINCT eventId) counter
  FROM matchEvent  
  GROUP BY eventId, playerId, matchId 
  HAVING SUM(eventName = 'goal') > 0 AND SUM(description = 'ownGoal') = 0
) t
GROUP BY playerId, matchId

见演示。

或者：

SELECT playerId, matchId, COUNT(*) total_goals
FROM (
  SELECT playerId, matchId
  FROM matchEvent  
  GROUP BY eventId, playerId, matchId 
  HAVING SUM(eventName = 'goal') > 0 AND SUM(description = 'ownGoal') = 0
) t
GROUP BY playerId, matchId

见演示。

或者：

SELECT playerId, matchId, COUNT(DISTINCT eventId) total_goals
FROM matchEvent  
WHERE eventName = 'goal' 
  AND eventId NOT IN (SELECT eventId FROM matchEvent WHERE description = 'ownGoal')
GROUP BY playerId, matchId

见演示。

结果：

| playerId | matchId | counter |
| -------- | ------- | ------- |
| 50       | 1       | 1       |