MySQL 查询的联接表中的 GROUP BY，COUNT()

Question

我希望在一组两个日期内按类别统计我的所有项目。我有一个主 table:

Table 项目

+-----+-------------+------------------+
| uid |    name     | project_category |
+-----+-------------+------------------+
|   1 | Testproject |                1 |
|   2 | Anothertest |                2 |
|   3 | Fietspomp   |                1 |
|   4 | Caramba     |                1 |
+-----+-------------+------------------+

谁的类别存储在project_category

+-----+-------------+
| uid |    name     |
+-----+-------------+
|   1 | Automotive  |
|   2 | Hospitality |
|   3 | Other       |
+-----+-------------+

不过，我的日期存储在以下 table 中。每个项目可以有更多日期（班次）。我只想找到每个项目的最早日期（班次）：我需要过滤每个项目的开始日期（= 第一个班次）。

Table project_shift

+-----+-------------+------------+
| uid | project_uid |    date    |
+-----+-------------+------------+
|   1 |           1 | 2015-05-03 |
|   2 |           2 | 2015-05-02 |
|   3 |           2 | 2015-06-04 |
|   4 |           1 | 2015-03-08 |
|   5 |           1 | 2015-08-08 |
+-----+-------------+------------+

我一直在寻找高低，但我的计数总是乱七八糟。这是我试过的：

SELECT pc.name as category, p.name, count(p.uid)
FROM project p
INNER JOIN project_category pc ON pc.uid = p.project_category
INNER JOIN project_shift ps    ON ps.project_uid = p.uid AND ps.date BETWEEN "2015-01-01" AND "2015-08-08"
GROUP BY pc.uid

另外：

SELECT    pc.name as category, count(pc.uid) as amount
FROM      project p
LEFT JOIN project_category pc ON pc.uid = p.project_category
WHERE     (SELECT MIN(date) FROM project_shift ps WHERE ps.project_uid = p.uid LIMIT 1) BETWEEN :date_from AND :date_to
GROUP BY  pc.uid

最后一个项目很接近，但仍然多了 29 个项目。

我在这里做错了什么？

编辑：所需结果如下（虚拟数据）

+-------------+--------+
|  category   | amount |
+-------------+--------+
| Automotive  |     70 |
| Hospitality |     22 |
| Other       |      2 |
+-------------+--------+

Answer 1

试试这个：

SELECT    pc.name as category, count(pc.uid) as amount
FROM      project p
INNER JOIN project_category pc ON pc.uid = p.project_category
INNER JOIN
    (
        SELECT MIN(`date`) AS `date`, project_uid FROM project_shift 
        WHERE `date` BETWEEN :date_from AND :date_to
        GROUP BY project_uid
    ) AS ps ON (ps.project_uid = pc.uid) 
GROUP BY  pc.uid

Answer 2

您还需要将项目 table 中的 project_category 添加到组中：

SELECT pc.name as category,  count(p.project_category)
FROM project p, project_category pc, project_shift ps
where pc.uid = p.project_category
and ps.project_uid = p.uid
and ps.date BETWEEN '1/1/2015' and '8/8/2015'
group by pc.uid, p.project_category

应该做的。

Answer 3

您的查询似乎基本正确。只是，project_category 您应该从中进行选择，并且应该进行外部连接的项目（例如获取项目为零的类别）。因此，您要按类别计算项目（即 p.uid）。

那么你的 LIMIT 子句就没有意义了。 MIN(date) 为您提供一条记录，其中包含项目的最短日期。这里不需要 LIMIT 子句。

SELECT    pc.name as category, count(p.uid) as amount
FROM      project_category pc 
LEFT JOIN project p ON p.project_category = pc.uid
WHERE     (SELECT MIN(date) FROM project_shift ps WHERE ps.project_uid = p.uid)
            BETWEEN :date_from AND :date_to
GROUP BY  pc.uid;

MySQL 查询的联接表中的 GROUP BY，COUNT()

GROUP BY, COUNT() in joined tables of MySQL query

mysql

join

group-by

count