如何在 SQL 中正确使用 GROUP BY 语句？答案

【问题标题】：How to properly using a GROUP BY statement in SQL?如何在 SQL 中正确使用 GROUP BY 语句？
【发布时间】：2019-03-27 06:19:50
【问题描述】：

我需要对潜在客户数据库运行查询，该数据库需要细分商店名称、该商店的潜在客户总数以及转换后的潜在客户总数。到目前为止，以下是我得到的：

SELECT  stores.store_name as Store, COUNT(*) as "Total Lead", 
        (SELECT COUNT(*) as "Converted Lead" FROM leads WHERE lead_status_id = 5) 
FROM "leads" 
INNER JOIN "employee_leads" ON "employee_leads"."lead_id" = "leads"."id" 
INNER JOIN "employees" ON "employees"."id" = "employee_leads"."employee_id" 
INNER JOIN "emp_stores" ON "emp_stores"."employee_id" = "employees"."id" 
INNER JOIN "stores" ON "stores"."id" = "emp_stores"."store_id" 
GROUP BY store_name;

但是，这是错误的，因为转换的部分计算了所有转换的潜在客户，而不是某个商店。我得到的结果如下：

 +------------+-------------+----------------+
 | Store      | Total Lead  | Converted Lead |
 +------------+-------------+----------------+
 | Store1     |         10  |            10  |
 | Store2     |         14  |            10  |
 | Store3     |         17  |            10  |
 +------------+-------------+----------------+

目前在“已转化的潜在客户”列中列出了所有商店中已转化的潜在客户总数。 “Converted Lead”列假设 Store1 为 4，Store2 为 3，Store3 为 4。我试图做 sub-groupby，但是，我得到了一个错误，因为每个商店会有超过 1 行。我需要查询计算每个商店的转换潜在客户，如下所示：

 +------------+-------------+----------------+
 | Store      | Total Lead  | Converted Lead |
 +------------+-------------+----------------+
 | Store1     |         10  |             4  |
 | Store2     |         14  |             3  |
 | Store3     |         17  |             4  |
 +------------+-------------+----------------+

有人能指出我正确的方向吗？谢谢

【问题讨论】：

标签： mysql sql

【解决方案1】：

您可以尝试使用条件聚合函数代替子查询。

SELECT  stores.store_name as Store, COUNT(*) as "Total Lead", 
        COUNT(CASE WHEN lead_status_id = 5 THEN 1 END) "Converted Lead" 
FROM "leads" 
INNER JOIN "employee_leads" ON "employee_leads"."lead_id" = "leads"."id" 
INNER JOIN "employees" ON "employees"."id" = "employee_leads"."employee_id" 
INNER JOIN "emp_stores" ON "emp_stores"."employee_id" = "employees"."id" 
INNER JOIN "stores" ON "stores"."id" = "emp_stores"."store_id" 
GROUP BY store_name;

如果你使用mysql，你可以尝试使用COUNT(lead_status_id = 5)

SELECT  stores.store_name as Store, COUNT(*) as "Total Lead", 
        COUNT(lead_status_id = 5) "Converted Lead" 
FROM "leads" 
INNER JOIN "employee_leads" ON "employee_leads"."lead_id" = "leads"."id" 
INNER JOIN "employees" ON "employees"."id" = "employee_leads"."employee_id" 
INNER JOIN "emp_stores" ON "emp_stores"."employee_id" = "employees"."id" 
INNER JOIN "stores" ON "stores"."id" = "emp_stores"."store_id" 
GROUP BY store_name;

【讨论】：

【解决方案2】：

您没有将子查询与商店名称相关联。虽然这可以在其 where 子句中完成，但更简单的方法是计算 case 表达式以仅计算具有相关状态的潜在客户：

SELECT  stores.store_name as Store, 
        COUNT(*) as "Total Lead", 
        COUNT(CASE lead_status_id WHEN 5 THEN 1 END) AS "Converted Lead"
FROM "leads" 
INNER JOIN "employee_leads" ON "employee_leads"."lead_id" = "leads"."id" 
INNER JOIN "employees" ON "employees"."id" = "employee_leads"."employee_id" 
INNER JOIN "emp_stores" ON "emp_stores"."employee_id" = "employees"."id" 
INNER JOIN "stores" ON "stores"."id" = "emp_stores"."store_id" 
GROUP BY store_name;

【讨论】：

【解决方案3】：

请试试这个。

SELECT  stores.store_name as Store, COUNT(*) as "Total Lead", 
        (SELECT COUNT(*) as "Converted Lead" FROM leads AS SS WHERE SS.id =A.id AND lead_status_id = 5) 
FROM "leads" AS A 
INNER JOIN "employee_leads" ON "employee_leads"."lead_id" = "A"."id" 
INNER JOIN "employees" ON "employees"."id" = "employee_leads"."employee_id" 
INNER JOIN "emp_stores" ON "emp_stores"."employee_id" = "employees"."id" 
INNER JOIN "stores" ON "stores"."id" = "emp_stores"."store_id" 
GROUP BY store_name;

【讨论】：