MySQL的计数和基于2列

Question

如果我有表answers以下数据分组结果：

+-------------------------------------------+ 
| forum_id | user_id | question_id | answer | 
+-------------------------------------------+ 
| 1  | 1  | 1   | A  | 
| 1  | 1  | 2   | B  | 
| 1  | 2  | 1   | B  | 
| 1  | 2  | 2   | B  | 
| 1  | 3  | 1   | A  | 
| 1  | 3  | 2   | B  | 
| 2  | 1  | 1   | A  | 
| 2  | 2  | 2   | A  | 
+-------------------------------------------+ 

我将如何选择答案的理货？重演结果如：

+------------------------------+ 
| question_id | answer | count | 
+------------------------------+ 
| 1   | A  | 2  | 
| 1   | B  | 1  | 
| 2   | A  | 3  | 
| 2   | B  | 0  | 
+------------------------------+ 

这是我目前所拥有的。我知道它必须使用GROUP BY，但我正在努力返回预期的结果。

SELECT question_id, answer, COUNT(*) 
FROM answers 
WHERE forum_id = 1 
GROUP BY question_id, answer 

Answer 1

试试这个：

SELECT t.question_id, t.answer, COUNT(DISTINCT user_id) 
FROM 
(
    SELECT DISTINCT a.question_id, t.answer 
    FROM answers as a, (SELECT 'A' AS answer UNION ALL SELECT 'B') AS t 
) AS t 
LEFT JOIN answers AS a ON a.question_id = t.question_id 
AND a.answer = t.answer and forum_id = 1 
GROUP BY t.question_id, t.answer; 

• demo

结果：

| question_id | answer | count | 
|-------------|--------|-------| 
|   1 |  A |  2 | 
|   1 |  B |  1 | 
|   2 |  A |  0 | 
|   2 |  B |  3 | 

需要注意的是：

在您的样本数据B中有3个问题2的条目，而A没有问题2的条目。而当然，所有论坛= 1，所以您的预期结果是不正确的，我认为我的是正确的。