HiveSQL——设计一张最近180天的注册、活跃留存表

0 问题描述

现有一个用户活跃表user_active(user_id,active_date)、用户注册表user_regist(user_id,regist_date)，表中分区字段都为dt(yyyy-MM-dd)，用户字段均为user_id; 设计一张 1-180天的注册活跃留存表；表结构如下：

1 数据分析

完整的代码如下：

selectregist_date,diff,active_user_cnt,casewhen nvl(regis_cnt, 0) != 0 then active_user_cnt / regis_cntend as retention_rate
from (selectt1.regist_date,max(t1.regist_count)                     as regis_cnt,datediff(t2.active_date, t1.regist_date) as diff,count(t2.user_id)                        as active_user_cntfrom (selectuser_id,to_date(regist_date)                                    as regist_date,count(user_id) over (partition by to_date(regist_date)) as regist_countfrom user_registwhere dt >= date_sub(current_date(), 180)) t1left join(selectuser_id,to_date(active_date) as active_datefrom user_activewhere dt >= date_sub(current_date(), 180)group by user_id, to_date(active_date)) t2on t1.user_id = t2.user_idwhere datediff(active_date, regist_date) >= 1and datediff(active_date, regist_date) <= 180group by t1.regist_date, datediff(t2.active_date, t1.regist_date)) t3
order by regist_date,diff;

上述代码解析：

步骤一：基于注册表，求出用户的注册日期regist_date、每日的用户注册数量regist_count

selectuser_id,to_date(regist_date)                                    as regist_date,count(user_id) over (partition by to_date(regist_date)) as regist_count
from user_regist
where dt >= date_sub(current_date(), 180);

步骤二：将用户注册表作为主表，关联活跃表，关联键为user_id，一对多的关系，形成笛卡尔积。需要注意：活跃用户表，每个用户每天可能会有多次活跃的情况，因此需要去重。

selectt1.regist_date,t1.user_id,t1.regist_count,t2.user_id,t2.active_date,datediff(t2.active_date, t1.regist_date) as diff
from (selectuser_id,to_date(regist_date)                                    as regist_date,count(user_id) over (partition by to_date(regist_date)) as regist_countfrom user_registwhere dt >= date_sub(current_date(), 180)) t1left join(selectuser_id,to_date(active_date) as active_datefrom user_activewhere dt >= date_sub(current_date(), 180)group by user_id, to_date(active_date)) t2on t1.user_id = t2.user_id;

步骤三：基于注册日期，留存周期分组（以“天”为单位），计算该留存周期下的活跃用户数

selectt1.regist_date,max(t1.regist_count)                     as regis_cnt,datediff(t2.active_date, t1.regist_date) as diff,count(t2.user_id)                        as active_user_cntfrom (selectuser_id,to_date(regist_date)                                    as regist_date,count(user_id) over (partition by to_date(regist_date)) as regist_countfrom user_registwhere dt >= date_sub(current_date(), 180)) t1left join(selectuser_id,to_date(active_date) as active_datefrom user_activewhere dt >= date_sub(current_date(), 180)group by user_id, to_date(active_date)) t2on t1.user_id = t2.user_id
where datediff(active_date, regist_date) >= 1and datediff(active_date, regist_date) <= 180
group by t1.regist_date, datediff(t2.active_date, t1.regist_date);

步骤四：计算留存率retention_rate

selectregist_date,diff,active_user_cnt,casewhen nvl(regis_cnt, 0) != 0 then active_user_cnt / regis_cntend as retention_rate
from (selectt1.regist_date,max(t1.regist_count)                     as regis_cnt,datediff(t2.active_date, t1.regist_date) as diff,count(t2.user_id)                        as active_user_cntfrom (selectuser_id,to_date(regist_date)                                    as regist_date,count(user_id) over (partition by to_date(regist_date)) as regist_countfrom user_registwhere dt >= date_sub(current_date(), 180)) t1left join(selectuser_id,to_date(active_date) as active_datefrom user_activewhere dt >= date_sub(current_date(), 180)group by user_id, to_date(active_date)) t2on t1.user_id = t2.user_idwhere datediff(active_date, regist_date) >= 1and datediff(active_date, regist_date) <= 180group by t1.regist_date, datediff(t2.active_date, t1.regist_date)) t3
order by regist_date,diff;