Question

我有三张桌子：

区域表：

+--------------------------+---------------------+------+-----+---------+----------------+
| Field                    | Type                | Null | Key | Default | Extra          |
+--------------------------+---------------------+------+-----+---------+----------------+
| id                       | bigint(20) unsigned | NO   | PRI | NULL    | auto_increment |
| location_id              | bigint(20) unsigned | NO   | MUL | NULL    |                |
| impressions_count        | bigint(20) unsigned | YES  |     | 0       |                |
| created                  | datetime            | YES  |     | NULL    |                |
| modified                 | datetime            | YES  |     | NULL    |                |
+--------------------------+---------------------+------+-----+---------+----------------+

地点表：

+----------------+---------------------+------+-----+---------+----------------+
| Field          | Type                | Null | Key | Default | Extra          |
+----------------+---------------------+------+-----+---------+----------------+
| id             | bigint(20) unsigned | NO   | PRI | NULL    | auto_increment |
| retailer_id    | bigint(20) unsigned | NO   | MUL | NULL    |                |
| zones_count    | int(10) unsigned    | YES  |     | 0       |                |
| contacts_count | int(10) unsigned    | YES  |     | 0       |                |
| created        | datetime            | YES  |     | NULL    |                |
| modified       | datetime            | YES  |     | NULL    |                |
+----------------+---------------------+------+-----+---------+----------------+

零售商表：

+---------------------+---------------------+------+-----+---------+----------------+
| Field               | Type                | Null | Key | Default | Extra          |
+---------------------+---------------------+------+-----+---------+----------------+
| id                  | bigint(20) unsigned | NO   | PRI | NULL    | auto_increment |
| account_rep_id      | bigint(20) unsigned | YES  | MUL | NULL    |                |
| leadsource_id       | bigint(20) unsigned | YES  | MUL | NULL    |                |
| industry_id         | bigint(20)          | NO   | MUL | NULL    |                |
| name                | varchar(100)        | NO   |     | NULL    |                |
| locations_count     | int(10) unsigned    | YES  |     | 0       |                |
| created             | datetime            | YES  |     | NULL    |                |
| modified            | datetime            | YES  |     | NULL    |                |
+---------------------+---------------------+------+-----+---------+----------------+

我想做什么：

我正在从Retailers表中选择所有记录。 Retailers表在Locations中有许多记录，而这些记录在Zones中有很多记录。

Zones表有一个名为impressions_count的字段。我想要做的是，在加入Retailers表和Locations表时，从Zones表中选择所有记录。

基本上，我认为我需要为Locations中的每个匹配记录返回一个SUM（Zones.impressions_count），然后为Retailers中的每个记录返回一个总和。我一直在靠墙撞墙 - 并且感谢任何指导！

最终：我的结果集应如下所示：

**from retailers**
id,
account_rep_id,
lead_source_id,
industry_id,
impressions_count,   // <- sum of related records in `Zones`

Answer 1

当您加入到位置时，您可能希望使用内联视图在加入之前按位置聚合数据......所以......

这样做是因为表之间的一对多关系不会通过Location_ID人为地增加impression_count的总和

Select * 
FROM retailers R
INNER JOIN Locations L
  on R.ID = L.retailer_id
INNER JOIN 
(Select sum(impressions_count) as cnt, Location_ID from Zones group by Location_ID) as Z
  on Z.Location_ID = L.ID

现在..如果你没有在你的输出中显示位置而你想要零售商的总和..那么我们需要再次总结..这次是零售商对每个地点的价值总结......

Select R.id
      ,R.account_rep_id
      ,R.lead_source_id
      ,R.industry_id
      ,coalesce(sum(cnt),0) as impressions_count
FROM retailers R
LEFT JOIN Locations L
  on R.ID = L.retailer_id
LEFT JOIN 
(Select coalesce(sum(impressions_count),0) as cnt, Location_ID from Zones group by Location_ID) as Z
  on Z.Location_ID = L.ID
GROUP BY R.id
        ,R.account_rep_id
        ,R.lead_source_id
        ,R.industry_id

如何获得相关表中字段的计数总和？

1 个答案: