【項目】數倉項目（十）

（圖片來源於網絡，侵刪）

---

一、數倉搭建 - ADS 層

1.1 設備主題

1.1.1 活躍設備數（日、周、月）

需求定義：

日活：當日活躍的設備數

周活：當週活躍的設備數

月活：當月活躍的設備數

1）建表語句

drop table if exists ads_uv_count;
create external table ads_uv_count(
`dt` string COMMENT '統計日期',
`day_count` bigint COMMENT '當日用戶數量',
`wk_count` bigint COMMENT '當週用戶數量',
`mn_count` bigint COMMENT '當月用戶數量',
`is_weekend` string COMMENT 'Y,N 是否是週末,用於得到本週最終結果',
`is_monthend` string COMMENT 'Y,N 是否是月末,用於得到本月最終結果'
) COMMENT '活躍設備數'
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_uv_count/';

2）導入數據

insert into table ads_uv_count
select
'2020-03-10' dt,
daycount.ct,
wkcount.ct,
mncount.ct,
if(date_add(next_day('2020-03-10','MO'),-1)='2020-03-10','Y','N') ,
if(last_day('2020-03-10')='2020-03-10','Y','N')
from
(
select
'2020-03-10' dt,
count(*) ct
from dwt_uv_topic
where login_date_last='2020-03-10'
)daycount join
(
select
'2020-03-10' dt,
count (*) ct
from dwt_uv_topic
where login_date_last>=date_add(next_day('2020-03-10','MO'),-7)
and login_date_last<= date_add(next_day('2020-03-10','MO'),-1)
) wkcount on daycount.dt=wkcount.dt
join
(
select
'2020-03-10' dt,
count (*) ct
from dwt_uv_topic
where
date_format(login_date_last,'yyyy-MM')=date_format('2020-03-10','yyyy-MM')
)mncount on daycount.dt=mncount.dt;

3）查詢導入結果

select * from ads_uv_count;

1.1.2 每日新增設備

1）建表語句

drop table if exists ads_new_mid_count;
create external table ads_new_mid_count
(
`create_date` string comment '創建時間' ,
`new_mid_count` BIGINT comment '新增設備數量'
) COMMENT '每日新增設備信息數量'
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_new_mid_count/';

2）導入數據

insert into table ads_new_mid_count
select
login_date_first,
count(*)
from dwt_uv_topic
where login_date_first='2020-03-10'
group by login_date_first;

3）查詢導入數據

select * from ads_new_mid_count;

1.1.3 沉默用戶數

需求定義：

沉默用戶：只在安裝當天啓動過，且啓動時間是在 7 天前

1）建表語句

drop table if exists ads_silent_count;
create external table ads_silent_count(
`dt` string COMMENT '統計日期',
`silent_count` bigint COMMENT '沉默設備數'
)
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_silent_count';

2）導入 2020-03-20 數據

insert into table ads_silent_count
select
'2020-03-15',
count(*)
from dwt_uv_topic
where login_date_first=login_date_last
and login_date_last<=date_add('2020-03-15',-7);

3）查詢導入數據

select * from ads_silent_count;

1.1.4 本週迴流用戶數

需求定義：

本週迴流用戶：上週未活躍，本週活躍的設備，且不是本週新增設備

1）建表語句

drop table if exists ads_back_count;
create external table ads_back_count(
`dt` string COMMENT '統計日期',
`wk_dt` string COMMENT '統計日期所在周',
`wastage_count` bigint COMMENT '迴流設備數'
)
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_back_count';

2）導入數據：

insert into table ads_back_count
select
'2020-03-15',
count(*)
from
(
select
mid_id
from dwt_uv_topic
where login_date_last>=date_add(next_day('2020-03-15','MO'),-7)
and login_date_last<= date_add(next_day('2020-03-15','MO'),-1)
and login_date_first<date_add(next_day('2020-03-15','MO'),-7)
)current_wk
left join
(
select
mid_id
from dws_uv_detail_daycount
where dt>=date_add(next_day('2020-03-15','MO'),-7*2)
and dt<= date_add(next_day('2020-03-15','MO'),-7-1)
group by mid_id
)last_wk
on current_wk.mid_id=last_wk.mid_id
where last_wk.mid_id is null;

3）查詢結果

select * from ads_back_count;

1.1.5 流失用戶數

需求定義：

流失用戶：最近 7 天未活躍的設備

1）建表語句

drop table if exists ads_wastage_count;
create external table ads_wastage_count(
`dt` string COMMENT '統計日期',
`wastage_count` bigint COMMENT '流失設備數'
)
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_wastage_count';

2）導入 2020-03-20 數據

insert into table ads_wastage_count
select
'2020-03-20',
count(*)
from
(
select
mid_id
from dwt_uv_topic
where login_date_last<=date_add('2020-03-20',-7)
group by mid_id
)t1;

3）查詢結果

select * from ads_wastage_count;

1.1.6 留存率

1）建表語句

drop table if exists ads_user_retention_day_rate;
create external table ads_user_retention_day_rate
(
`stat_date` string comment '統計日期',
`create_date` string comment '設備新增日期',
`retention_day` int comment '截止當前日期留存天數',
`retention_count` bigint comment '留存數量',
`new_mid_count` bigint comment '設備新增數量',
`retention_ratio` decimal(10,2) comment '留存率'
) COMMENT '每日用戶留存情況'
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_user_retention_day_rate/';

2）導入數據

insert into table ads_user_retention_day_rate
select
'2020-03-10',--統計日期
date_add('2020-03-10',-1),--新增日期
1,--留存天數
sum(if(login_date_first=date_add('2020-03-10',-1) and
login_date_last='2020-03-10',1,0)),--2020-03-09 的 1 日留存數
sum(if(login_date_first=date_add('2020-03-10',-1),1,0)),--2020-03-09 新增
sum(if(login_date_first=date_add('2020-03-10',-1) and
login_date_last='2020-03-10',1,0))/sum(if(login_date_first=date_add('2020-03-10',-
1),1,0))*100
from dwt_uv_topic
union all
select
'2020-03-10',--統計日期
date_add('2020-03-10',-2),--新增日期
2,--留存天數
sum(if(login_date_first=date_add('2020-03-10',-2) and
login_date_last='2020-03-10',1,0)),--2020-03-08 的 2 日留存數
sum(if(login_date_first=date_add('2020-03-10',-2),1,0)),--2020-03-08 新增
sum(if(login_date_first=date_add('2020-03-10',-2) and
login_date_last='2020-03-10',1,0))/sum(if(login_date_first=date_add('2020-03-10',-
2),1,0))*100
from dwt_uv_topic
union all
select
'2020-03-10',--統計日期
date_add('2020-03-10',-3),--新增日期
3,--留存天數
sum(if(login_date_first=date_add('2020-03-10',-3) and
login_date_last='2020-03-10',1,0)),--2020-03-07 的 3 日留存數
sum(if(login_date_first=date_add('2020-03-10',-3),1,0)),--2020-03-07 新增
sum(if(login_date_first=date_add('2020-03-10',-3) and
login_date_last='2020-03-10',1,0))/sum(if(login_date_first=date_add('2020-03-10',-
3),1,0))*100
from dwt_uv_topic;

3）查詢導入數據

select * from ads_user_retention_day_rate;

1.1.7 最近連續三週活躍用戶數

1）建表語句

drop table if exists ads_continuity_wk_count;
create external table ads_continuity_wk_count(
`dt` string COMMENT '統計日期,一般用結束週週日日期,如果每天計算一次,可用當天日
期',
`wk_dt` string COMMENT '持續時間',
`continuity_count` bigint COMMENT '活躍次數'
)
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_continuity_wk_count';

2）導入 2020-03-20 所在周的數據

insert into table ads_continuity_wk_count
select
'2020-03-15',
concat(date_add(next_day('2020-03-15','MO'),-7*3),'_',date_add(next_day('
2020-03-15','MO'),-1)),
count(*)
from
(
select
mid_id
from
(
select
mid_id
from dws_uv_detail_daycount
where dt>=date_add(next_day('2020-03-10','monday'),-7)
and dt<=date_add(next_day('2020-03-10','monday'),-1)
group by mid_id
union all
select
mid_id
from dws_uv_detail_daycount
where dt>=date_add(next_day('2020-03-10','monday'),-7*2)
and dt<=date_add(next_day('2020-03-10','monday'),-7-1)
group by mid_id
union all
select
mid_id
from dws_uv_detail_daycount
where dt>=date_add(next_day('2020-03-10','monday'),-7*3)
and dt<=date_add(next_day('2020-03-10','monday'),-7*2-1)
group by mid_id
)t1
group by mid_id
having count(*)=3
)t2

3）查詢

select * from ads_continuity_wk_count;

1.1.8 最近七天內連續三天活躍用戶數

1）建表語句

drop table if exists ads_continuity_uv_count;
create external table ads_continuity_uv_count(
`dt` string COMMENT '統計日期',
`wk_dt` string COMMENT '最近 7 天日期',
`continuity_count` bigint
) COMMENT '連續活躍設備數'
row format delimited fields terminated by '\t'
location '/warehouse/gmall/ads/ads_continuity_uv_count';

2）寫出導入數據的 SQL 語句

insert into table ads_continuity_uv_count
select
'2020-03-12',
concat(date_add('2020-03-12',-6),'_','2020-03-12'),
count(*)
from
(
select mid_id
from
(
select mid_id
from
(
select
mid_id,
date_sub(dt,rank) date_dif
from
(
select
mid_id,
dt,
rank() over(partition by mid_id order by dt) rank
from dws_uv_detail_daycount
where dt>=date_add('2020-03-12',-6) and
dt<='2020-03-12'
)t1
)t2
group by mid_id,date_dif
having count(*)>=3
)t3
group by mid_id
)t4;

3）查詢

select * from ads_continuity_uv_count;

---

結束語

---

都看到這裏了，點贊評論一下吧！！！

點擊查看👇

敬請期待！