MySQL：从按类别组织的两列中获取总计

考虑MYSQL数据库中的下表：

表1：

+------+-----------+----------------------+
| id   |   atual   | category_id  |  user |
+------+-----------+--------------|-------+
| 1    |   100     |  1           |   1   |
| 2    |   150     |  2           |   1   |
| 3    |   50      |  1           |   2   |
+------+-----------+--------------|-------+

表2：

+------+-----------+----------------------+
| id   |   budget  | category_id  |  user |
+------+-----------+--------------|-------+
| 1    |   100     |  2           |   1   |
| 2    |   150     |  1           |   2   |
| 3    |   50      |  1           |   1   |
+------+-----------+--------------|-------+

表3：

+------+-----------+
| id   |   name    | 
+------+-----------+
| 1    |   one     |
| 2    |   two     |
| 3    |   three   |
+------+-----------+

我想计算表1和表2中给定用户（在我的示例中为1）的“年度”和“预算”总计，按类别名称组织：

我尝试了以下查询，该查询给出了不分类别的年度和预算总额：

SELECT table2.id, table3.name AS name_category, SUM( budget ) , 
(SELECT SUM( atual) FROM table1 WHERE user =1)
FROM table2 INNER JOIN table3
ON table2.category_id=table3.id

以下是一种方法：

select t3.id, t3.name, sum(actual) as actual, sum(budget) as budget
from ((select category_id, sum(actual) as actual, NULL as budget
       from table1
       where user = 1
       group by category_id
      ) union all
      (select category_id, NULL as actual, sum(budget) as budget
       from table2
       where user = 1
       group by category_id
      )
     ) ab join
     table3 t3
     on ab.category_id = t3.id
group by t3.id, t3.name;

---

这个问题假设按类别组织，所以我认为内部查询应该有GROUP by category_id和SELECT部分中的“category_id”字段。这看起来很奇怪，但（几乎）起作用了：它正确地添加了按类别分组的实际值，但将所有预算值添加到了一个类别中（尽管有多个类别的预算）预算数字以总额的形式呈现，其中实际数字为：Null@NunoNogueira…问题已解决。我仍在

预算

子查询中选择

用户ID

。我需要几个小时来考虑这是如何完成的，但它工作正常，谢谢！