Question

我正在尝试根据我的销售表打印报告，但很难写出我的查询。

 +----------+--------+-----------+-------+------------+
 |                      SALES                        |
 +----------+--------+-----------+-------+------------+
 | agent_id |  type  |  result   | value |    date    |
 +----------+--------+-----------+-------+------------+
 |        1 | sales  | yes       | 12.34 | 2018-05-01 |
 |        1 | sales  | yes       |  5.45 | 2018-05-01 |
 |        1 | return | other     |  7.00 | 2018-05-01 |
 |        1 | return | saved     | 19.99 | 2018-05-01 |
 |        1 | return | not_saved |  1.19 | 2018-05-01 |
 |        2 | return | saved     |  1.34 | 2018-05-01 |
 |        2 | return | not_saved | 29.04 | 2018-05-01 |
 |        2 | sales  | yes       | 11.11 | 2018-05-01 |
 |        3 | sales  | no        | 10.00 | 2018-05-01 |
 |        3 | sales  | no        | 89.34 | 2018-05-01 |
 |        3 | sales  | yes       |  1.41 | 2018-05-01 |
 |        3 | sales  | other     |  2.41 | 2018-05-01 |
 |        3 | sales  | other     |  2.41 | 2018-04-28 |
 +----------+--------+-----------+-------+------------+

我在下面的查询中第3，第4和第5列遇到了困难。

COLUMN1 = agent_id

COLUMN2 =通话次数。这是每个代理的记录数

COLUMN3 =保存数量。这将记录type为'销售'并且result为'是'的记录

COLUMN4 =保存的值。这是总和（值），其中type =“销售”和result =“是”

COLUMN5 =转换率。这将是（column3 / column2）x100

这个查询是我能得到的。我无法弄清楚剩下的。任何人都可以指导我朝正确的方向发展吗？

SELECT agent_id,count(*) AS 'Number of Calls',COLUMN3,COLUMN4,COLUMN5 
FROM sales 
WHERE comm_ts > '2018-04-29' 
GROUP BY agent_id;

这将是我的结果集

 +----------+--------+----------+-------+------------+---------+------------+
 | agent_id |  Number of Calls  |     COLUMN3        | COLUMN4 |   COLUMN5  |
 +----------+--------+-----------+-------------------+---------+------------+
 |        1 |       5          |        2           | 17.79   |     40%    |
 |        2 |       3          |        1           |  11.11  |     33%    |
 |        3 |       4          |        1           |  1.41   |     25%    |
 +----------+--------+----------+--------------------+---------+------------+

Answer 1

我认为你想要条件聚合：

SELECT agent_id, count(*) AS num_calls,
       sum(case when type = 'sales' and result = 'yes' then 1 else 0 end) as column3,
       sum(case when type = 'sales' and result = 'yes' then value else 0 end) as column4,
       avg(case when type = 'sales' and result = 'yes' then 1.0 else 0 end) as column5
FROM sales 
WHERE comm_ts > '2018-04-29' 
GROUP BY agent_id;

在MySql中，如何基于一列进行分组并对其他列进行求和？

1 个答案: