淘宝用户分析

SELECT DISTINCT user_id,item_id,timestamp FROM userbehaviors;

CREATE TABLE userbehavior AS 
(SELECT user_id ,item_id,category_id,behavior_type,timestamp,
FROM_UNIXTIME(timestamp,‘%Y-%m-%d‘)AS Date,
FROM_UNIXTIME(timestamp,‘%H‘)AS Time FROM userbehavior)

SELECT MAX(Date),MIN(Date),MIN(Time),MAX(Time)
FROM userbehavior;

DELETE FROM userbehavior
WHERE Date<‘2017-11-25‘ OR Date>‘2017-12-03‘;

SELECT COUNT(DISTINCT user_id)AS‘活跃用户数‘
FROM userbehavior;

SELECT Date,COUNT(DISTINCT user_id)AS‘日活跃用户数‘
FROM userbehavior
GROUP BY Date;

select count(behavior_type) as 浏览量 from UserBehavior
where behavior_type=‘pv‘;

select date,count(behavior_type)as 访问量 from UserBehavior
where behavior_type=‘pv‘
group by date;

select a.time,a.访问量,b.访客数 from
(select time,count(behavior_type)as 访问量 from UserBehavior
where behavior_type=‘pv‘
group by time)a
inner join 
(select time,count(distinct user_id) as 访客数 from UserBehavior
group by time)b
on a.time=b.time
order by time asc;

SELECT COUNT(DISTINCT user_id)AS‘用户数‘
FROM userbehavior
WHERE Date in(‘2017-12-02‘,‘2017-12-03‘) AND behavior_type=‘pv‘;

SELECT category_id AS‘商品类目ID‘,COUNT(user_id)AS‘用户浏览次数‘,
--concat()连接%,*100加%，round()取两位小数，1157642
CONCAT(ROUND(COUNT(user_id)*100/950323,2),‘%‘) AS‘浏览占比‘
FROM userbehavior
WHERE Date in(‘2017-12-02‘,‘2017-12-03‘) AND behavior_type=‘pv‘
GROUP BY category_id
ORDER BY 用户浏览次数 DESC LIMIT 10;

select AVG(浏览占比) FROM 
(SELECT category_id AS‘商品类目ID‘,COUNT(user_id)AS‘用户浏览次数‘,
CONCAT(ROUND(COUNT(user_id)*100/950323,2),‘%‘) AS‘浏览占比‘
FROM userbehavior
WHERE Date in(‘2017-12-02‘,‘2017-12-03‘) AND behavior_type=‘pv‘
GROUP BY category_id
ORDER BY 用户浏览次数 DESC ) as b ;

select sum(用户浏览次数) as ‘用户浏览总次数‘ FROM 
(SELECT category_id AS‘商品类目ID‘,COUNT(user_id)AS‘用户浏览次数‘,
CONCAT(ROUND(COUNT(user_id)*100/950323,2),‘%‘) AS‘浏览占比‘
FROM userbehavior
WHERE Date in(‘2017-12-02‘,‘2017-12-03‘) AND behavior_type=‘pv‘
GROUP BY category_id
ORDER BY 用户浏览次数 DESC ) as b ;

SELECT Time AS‘时间‘,COUNT(*)AS‘时活跃量‘,COUNT(DISTINCT user_id)AS‘时活跃用户数‘
FROM userbehavior
GROUP BY Time
ORDER BY Time ASC;

SELECT Time AS‘时间‘, count(*) AS ‘成交量‘
FROM userbehavior
where behavior_type=‘buy‘ 
GROUP BY Time
ORDER BY Time ASC;

SELECT COUNT(DISTINCT user_id)AS‘UV‘,(SELECT COUNT(*)FROM userbehavior)AS‘PV‘,
((SELECT COUNT(*)FROM userbehavior)/(SELECT COUNT(DISTINCT user_id)))AS‘PV/UV‘
FROM userbehavior;

select date,count(behavior_type) as 成交量 from UserBehavior
where behavior_type=‘buy‘
group by date
order by date asc;

SELECT COUNT(user_id)AS‘浏览即退出的次数‘,COUNT(DISTINCT user_id)AS‘浏览即退出的用户数‘
FROM userbehavior
WHERE user_id NOT IN(SELECT DISTINCT user_id FROM userbehavior WHERE behavior_type=‘fav‘)
AND user_id NOT IN(SELECT DISTINCT user_id FROM userbehavior WHERE behavior_type=‘cart‘)
AND user_id NOT IN(SELECT DISTINCT user_id FROM userbehavior WHERE behavior_type=‘buy‘);

SELECT COUNT(user_id)AS‘浏览次数‘
FROM usebehavior
WHERE  behavior_type=‘pv‘;

SELECT behavior_type AS‘用户行为‘,COUNT(*)AS‘用户行为数‘
FROM userbehavior
GROUP BY behavior_type
ORDER BY 用户行为数 DESC;

-- 加入购物车的用户数量/加入购物车购买的用户数量=转化率
SELECT COUNT(DISTINCT a.user_id)AS‘加入购物车数量‘,COUNT(DISTINCT b.user_id)AS‘加入购物车后购买数量‘
FROM (SELECT DISTINCT user_id,item_id,category_id,timestamp
      FROM userbehavior
      WHERE behavior_type=‘cart‘)AS a
LEFT JOIN (SELECT DISTINCT user_id,item_id,category_id,timestamp
      FROM userbehavior
      WHERE behavior_type=‘buy‘)AS b
-- 连接条件是user_id,item_id,category_id均相等，说名是同一个用户加入同一件商品进入购物车，但是timestamp时间不同
ON a.user_id=b.user_id AND a.item_id=b.item_id AND a.category_id=b.category_id AND a.timestamp<b.timestamp;

select * from
-- 浏览商品用户数量
(select count(distinct user_id)as‘浏览商品用户数目‘ from UserBehavior where behavior_type=‘pv‘) as b,
-- 加入购物车用户数量
(SELECT * FROM data.conversion_buy) as a
)

select * from(
-- 浏览商品用户数量
(select count(distinct user_id)as‘浏览商品用户数目‘ from UserBehavior where behavior_type=‘pv‘) as c,
-- 收藏后购买的转化率
(SELECT COUNT(DISTINCT a.user_id)AS‘用户收藏数量‘,COUNT(DISTINCT b.user_id)AS‘收藏后购买数量‘
FROM (SELECT DISTINCT user_id,item_id,category_id,timestamp
      FROM userbehavior
      WHERE behavior_type=‘fav‘)AS a
LEFT JOIN (SELECT DISTINCT user_id,item_id,category_id,timestamp
      FROM userbehavior
      WHERE behavior_type=‘buy‘)AS b
ON a.user_id=b.user_id AND a.item_id=b.item_id AND a.category_id=b.category_id AND a.timestamp<b.timestamp)as d)