mysql 关联表的筛选分组取最大值

SELECT b2.id bid,b2.`start_date`,b2.`name`,d.* FROM 
(
    SELECT c.*,MAX(IFNULL(c.sdate,'0')) max_start FROM 
    (
        SELECT a.id,a.`update_time`,b.id bid,b.a_id,b.start_date sdate,b.end_date edate
        from agreement a left join bills b 
        on b.a_id = a.id and b.sell_user_id = a.sell_user_id
        where a.sell_user_id = '160000001'
        AND a.`status` = 2
        AND a.`update_time` >= 1559318400
        AND a.`update_time` <= 1567439999
        AND IFNULL(b.start_date,'0') < a.end_date
        ORDER BY bid DESC
    ) c GROUP BY c.id
) d LEFT JOIN bills b2 ON d.id = b2.a_id AND b2.start_date = d.max_start
HAVING IFNULL(b2.`start_date`,'0') < FROM_UNIXTIME(d.`update_time`, "%Y-%m-%d");

网上参考了一翻，还是觉得有点差异，
上面的查询语句可以简化为下面的模型：

select b2*,d.*
(
    select c.*,MAX(c.date) max_start FROM
    (
         select * from table_a a left join table_b b on a.aid = b.bid where a.id = '160000001' 
    ) c GROUP BY c.aid
) d left join table_b b2 ON d.id = b2.bid AND b2.date = d.max_start

关联表的分组取最大值的数据库语句

其他知识点：
为空替换为0，否则输出原值

IFNULL(c.sdate,'0')
``
时间戳替换为时间格式

FROM_UNIXTIME(d.update_time, "%Y-%m-%d")