Question

我已经搜索过，但我找不到我想要的东西。

我有两张桌子：

表1

|  d  |      text         |
| 101 | 'description 101' |
| 102 | 'description 102' |
| 103 | 'description 103' |

表2

| id |  d1 |  d2 |  d3 |  d4 |
| 01 | 104 | 242 | 102 | 222 |
| 02 | 423 | 553 |     |     |
| 03 | 832 | 142 | 102 |     |

等

我想要计算Table1中每个“d”用作表2中的d1，d2，d3和d4的次数。输出看起来像这样：

|  d  | count_d1 | count_d2 | count_d3 | count_d4 |
| 101 |   30032  |   108    |   5002   |   392    |
| 102 |    440   |  5330    |   24     |     5    |
| 103 |     0    |   309    |    2220  |    4     |

等

我确信有一些显而易见的东西，我只是没想到，但我现在已经看了一个多小时了，而且我迷失在一连串的连接和子查询中。

Answer 1

SELECT t1.d,
(
  SELECT COUNT(*) 
  FROM Table2 s2a 
  WHERE s2a.d1 = t1.d
) AS count_d1,
(
  SELECT COUNT(*) 
  FROM Table2 s2b
  WHERE s2b.d2 = t1.d
) AS count_d2,
(
  SELECT COUNT(*) 
  FROM Table2 s2c
  WHERE s2c.d3 = t1.d
) AS count_d3,
(
  SELECT COUNT(*) 
  FROM Table2 s2d 
  WHERE s2d.d4 = t1.d
) AS count_d4
FROM Table1 as t1

Answer 2

您必须将其拆分为4个查询（每列1个）

SELECT d, sum(Count_d1), sum(Count_d2), sum(Count_d3), sum(Count_d4) FROM (
SELECT d, count(1) AS Count_d1, null AS Count_d2, null AS Count_d3, null AS Count_d4 FROM Table1 LEFT JOIN Table2 on Table1.d=Table2.d1
UNION ALL
SELECT d, null AS Count_d1, count(1) AS Count_d1, null AS Count_d3, null AS Count_d4 FROM Table1 LEFT JOIN Table2 on Table1.d=Table2.d2
UNION ALL
SELECT d, null AS Count_d1, null AS Count_d2, count(1) AS Count_d3, null AS Count_d4 FROM Table1 LEFT JOIN Table2 on Table1.d=Table2.d3
UNION ALL
SELECT d, null AS Count_d1, null AS Count_d2, null AS Count_d3, count(1) AS Count_d4 FROM Table1 LEFT JOIN Table2 on Table1.d=Table2.d4
) GROUP BY d

Answer 3

您可以通过多个别名JOIN条件来完成此操作，如下所示：

SELECT 
    Table1.d,
    COUNT(d1.id) AS d1,
    COUNT(d2.id) AS d2,
    COUNT(d3.id) AS d3,
    COUNT(d4.id) AS d4
FROM 
    Table1
     LEFT JOIN 
    Table2 d1 ON 
        Table1.d = d1.D1
     LEFT JOIN 
    Table2 d2 ON 
        Table1.d = d2.D2
     LEFT JOIN 
    Table2 d3 ON 
        Table1.d = d3.D3
     LEFT JOIN 
    Table2 d4 ON 
        Table1.d = d4.D4
GROUP BY Table1.d

这将允许您避免对表的多个单值调用的开销，同时保留单个查询。

在多列SQL中计算一个值

3 个答案: