【SQL】电商网站各模块指标的统计以及分析

以下这些都是针对于电商网站的指标来分析业务的：以下这些指标都是使用SQL语句分析得出的结果，首先先创建一张表，然后将数据导入表中，字段根据分析的业务的需要而定，这里我的选择是我已经存在的表进行查询的，下面将给出所有查询的SQL语句

模块分析：
select * from web_log.weblog_etl where dateStr="20180709" limit 5;

1.统计每天每小时的PV数
select day,hour,count(request) as pvs from web_log.weblog_detail where dateStr="20180709" group by day,hour

2.统计每天的pv数
select dayStr,day,count(request) as pvs from web_log.weblog_detail where dateStr="20180709" group by dayStr,day

3.统计每月的pv数
select dayStr,month,count(request) as pvs from web_log.weblog_detail where dateStr="20180709" group by dayStr,month

4.统计不同终端维度的PV数
select dayStr,user_agent,count(request) as pvs from web_log.weblog_detail where dateStr="20180709" group by dayStr,user_agent

5.统计不同终端类型
select user_agent,count(user_agent) as iphone from web_log.weblog_detail where dateStr="20180709" group by user_agent

6.统计每个月每个终端的pv数
select dayStr,month,day,hour,user_agent,count(request) as pvs from web_log.weblog_detail where dateStr="20180709" group by dayStr,month,day,hour,user_agent

7.统计人均pv数
select dayStr,count(request)/count(distinct ip) from web_log.weblog_detail where dateStr="20180709" group by dayStr

8.统计不同来源的pv数

select
*
from
(
select t.ref_host,t.dayStr,t.hour,t.pv,row_number() over (partition by t.ref_host order by t.pv) as rnk
from
(
select
ref_host,dayStr,hour,count(request) as pv
from
web_log.weblog_detail
where
dateStr="20180709"
group by
ref_host,dayStr,hour
)t
)n
where n.rnk <4
---------------
select
*
from
(
select t.ref_host,t.pv,row_number() over (partition by t.ref_host order by t.pv) as rnk
from
(
select
ref_host,count(request) as pv
from
web_log.weblog_detail
where
dateStr="20180709"
group by
ref_host
)t
)n
where n.rnk <4

9.受访分析
9.1统计每日热门的受访页面的topN
9.1.1先统计每日每个网页访问次数
select dayStr,request,count(request) as counts from weblog_detail
where dateStr="20180709" group by dayStr,request

(2)row_number
select dayStr,request,count(request) as pv,row_number() over (partition by request order by request) from weblog_detail
where dateStr="20180709" group by dayStr,request

10.访客分析
10.1每个访客每小时的访问数(session个数)
select
ip,substring(s_time,12,2) as hour,count(distinct sessionId) as number
from web_log.weblog_pv
where
dateStr="20180709"
group by ip,substring(s_time,12,2)

10.2每个小时的ip访问数
select day,hour,count(ip) from web_log.weblog_detail where dateStr="20180709" group by day,hour

10.3每天的/每月的ip访问数
select month,day,count(ip) from web_log.weblog_detail where dateStr="20180709" group by month,day

10.4统计新增访客---需要维护历史访客表

11.visit模型分析
11.1单次访客统计（sessionId只有一个）
select
substring(s_time,9,2),ip,count(distinct sessionId) as number
from
web_log.weblog_pv
where
dateStr="20180709"
group by
substring(s_time,9,2),ip
having number=1

11.2回头访客个数（sessionId >1)
select
substring(s_time,9,2),ip,count(distinct sessionId) as number
from
web_log.weblog_pv
where
dateStr="20180709"
group by
substring(s_time,9,2),ip
having number>1

11.3人均访问频率(一天访问多少次) 总session个数/总人数
select
substring(s_time,9,2) as day,count(distinct sessionId)/count(distinct ip) as number
from
web_log.weblog_pv
where
dateStr="20180709"
group by
substring(s_time,9,2)

11.4用户平均访问页数
select sum(pageNum)/count(distinct ip) from web_log.weblog_visit where dayStr="20180709"

12pageview模型分析
12.1每个用户的访问时长
select
ip,sum(stayTime) as alltime
from
web_log.weblog_pv
where dateStr="20180709"
group by ip
12.2 用户的平均访问时长
select sum(stayTime)/count(distinct ip) as avg_time
from
web_log.weblog_pv
where
dateStr="20180709"

12.3访问页面的漏斗

create table tb_num1 as
select "step1" as step, count(*) as number from web_log.weblog_pv where dateStr="20180709" and request like "%/hadoop-mahout%"

union all
select "step2" as step, count(*) as number from web_log.weblog_pv where dateStr="20180709" and request like "%/hadoop-hive%"

union all
select "step3" as step, count(*) as number from web_log.weblog_pv where dateStr="20180709" and request like "%/hadoop-zookeeper%"

union all
select "step4" as step,count(*) as number from web_log.weblog_pv where dateStr="20180709" and request like "%/black-ip-list%"

create table tb_num as
select "step1" as step, count(*) as number from web_log.weblog_pv where dayStr="20180709" and request like "%/hadoop-mahout%"
union all
select "step2" as step, count(*) as number from web_log.weblog_pv where dayStr="20180709" and request like "%/hadoop-hive%"
union all
select "step3" as step, count(*) as number from web_log.weblog_pv where dayStr="20180709" and request like "%/hadoop-zookeeper%"
union all
select "step4" as step,count(*) as number from web_log.weblog_pv where dayStr="20180709" and request like "%/black-ip-list%"

select
*
from
tb_num

-----全链接笛卡尔积
select
*
from
(select tb_num.step as ak,tb_num.number as av from tb_num) a,
(select tb_num.step as bk,tb_num.number as bv from tb_num) b

-----漏斗

select
t.ak,t.av,t.bk,t.bv,t.av/t.bv as retain_rate
from
(
select
*
from
(select tb_num.step as ak,tb_num.number as av from tb_num) a,
(select tb_num.step as bk,tb_num.number as bv from tb_num) b
) t
where
t.ak='step1' and t.bk='step1' or
t.ak='step2' and t.bk='step1' or
t.ak='step3' and t.bk='step2' or
t.ak='step4' and t.bk='step3'

实例：不使用hive中的函数

公司有多个部门，统计出各个部门中按照工资排序前3的。（9条）
   name       deptNo        sal
   1-1       1           9000
   1-2       1           6000
   1-3       1           7000
   2-1       2           8000
   3-2       3           7600
   2-3       2           8800
   2-5       2           8800
   3-3       3           5500
   3-4       1           5000

SQL语句：

select deptno, name, sal
from emp e1
where
(
select count(1)
from emp e2
where e2.deptno=e1.deptno and e2.sal>=e1.sal
) <=3 /*这里的数值表示你想取前几名*/
order by deptno, sal desc;

【SQL】电商网站各模块指标的统计以及分析相关推荐

电商网站如何进行库存同步处理Redis+Lua
电商网站库存模块库存表包含了商品的sku,商品类型,商品款号,颜色,尺码,库存数,版本号,创建时间,修改时间. 商品类型,可根据商品分为,普通商品,赠品,内卖商品,预售商品等库存表结构 @Data ...
项目案例——电商网站
初学程序结构写的一个简单网站,使用了大量的循环语句,代码重复较多,因此参考度不大. 下面是完整代码: """ 用户登录的电商网站 """ # ...
电商网站购物车存储方式
大型电商网站在购物节的时候都会有很高的并发量,购物车模块无疑是访问最多的,那么要如何设计才能扛住并发量呢 1.Session方式: 优点:初期性能好,访问快,保存在服务端,相对比较安全缺点:对于大型 ...
人气流量通通要！电商网站商品页设计超全面指南
编者按:今天这篇文章读完后,同学们可以来做一个小练习,根据文章,把优设看做一个电商网站,文章看做商品,来分析一下哪篇文章是热门,有哪些突出的方法,文章里藏有哪些引导用户的细节等等,即学即用,不容易忘, ...
《Spark SQL大数据实例开发》9.2 综合案例实战——电商网站搜索排名统计
<Spark SQL大数据实例开发>9.2 综合案例实战--电商网站搜索排名统计 9.2.1 案例概述本节演示一个网站搜索综合案例:以京东为例,用户登录京东网站,在搜索栏中输入搜 ...
电商网站中用户评论模块剖析
用户评论是电商网站交互设计中的重要模块之一,用户体验的好坏将直接影响到产品的销售数量,乃至整站的访问量.网上购物,我们只能看到一张张商品图片和描述,对于产品的优劣无法准确判断,因此,已购买者的评论对于 ...
电商网站运营的 7 大关键指标
本文介绍电商网站用户运营转化的相关指标体系,通过对这些指标的统计.监测和分析,可以及时发现电商运营的问题,以便有效及时改进和优化,提升电商转化率和销售额. 其中,不同类别指标对应电商运营的不同环节,如 ...
电商总结（八）如何打造一个小而精的电商网站架构
前面写过一些电商网站相关的文章,这几天有时间,就把之前写得网站架构相关的文章,总结整理一下.把以前的一些内容就连贯起来,这样也能系统的知道,一个最小的电商平台是怎么一步步搭建起来的.对以前的文章感兴趣 ...
用户在电商网站中购买成功了，那么 TA 在微服务中经历了什么？
题目:用户在电商网站中购买成功了,那么它在微服务中经历了什么? 当我傻啊,用户在电商网站购买成功,还在微服务中,那肯定就是有一套微服务架构的电商系统. 设计一套电商系统还不简单简单想象一下,既然是一 ...

【SQL】电商网站各模块指标的统计以及分析

【SQL】电商网站各模块指标的统计以及分析相关推荐

最新文章

热门文章