具有列值总和的MySQL Pivot查询

2019年7月28日 147次阅读

表1(报告) – 数据结构

id  distributorId clientId amount
1       1            162    2500
2       2            163    3500
3       3            203    4500
4       6            157    5500
5       1            162    1500
6       3            163    2000
7       3            162    1000

表2(经销商) – 带数据的结构

id    distributor
1        Dis A
2        Dis B
3        Dis C
6        Dis D

表3(客户端) – 具有数据的结构

id    client_name
162     Client A
163     Client B
203     Client C
157     Client D

期望的输出使用上面定义的3个表：

client_name    Dis A    Dis B    Dis C    Dis D
 Client A      4000      0       1000      0
 Client B       0        3500    2000      0
 Client C       0        0       4500      0
 Client D       0        0        0       5500

最佳答案 对于fixed pivot sql,试试这个：

select
    t.client_name,
    sum(if(t.distributor = 'Dis A', t.amount, 0)) as `Dis A`,
    sum(if(t.distributor = 'Dis B', t.amount, 0)) as `Dis B`,
    sum(if(t.distributor = 'Dis C', t.amount, 0)) as `Dis C`,
    sum(if(t.distributor = 'Dis D', t.amount, 0)) as `Dis D`
from (
    select c.id, c.client_name, d.distributor, r.amount
    from clients c
    left join reports r on c.id = r.clientId
    left join distributor d on d.id = r.distributorId
)t
group by t.id
order by t.client_name

SQLFiddle DEMO HERE

对于动态pivot sql,试试这个：

SET @sql = NULL;
SELECT
  GROUP_CONCAT(DISTINCT
    CONCAT(
      'sum(if(t.distributor = ''',
      distributor,
      ''', t.amount, 0)) AS `',
      distributor ,'`'
    )
  ) INTO @sql
FROM distributor;
SET @sql = CONCAT('SELECT t.client_name, ', @sql, ' FROM (
    select c.id, c.client_name, d.distributor, r.amount
    from clients c
    left join reports r on c.id = r.clientId
    left join distributor d on d.id = r.distributorId
)t
group by t.id
order by t.client_name');

PREPARE stmt FROM @sql;
EXECUTE stmt;
DEALLOCATE PREPARE stmt;

SQLFiddle DEMO HERE

点赞