hive的排序，分組练习

                                            hive的排序，分組练习

数据：

添加表和插入数据(数据在Linux本地中)

create table if not exists tab1(
  IP        string,
  SOURCE    string,
  TYPE      string
)
row format delimited fields terminated by '|'
stored as textfile;

load data local inpath '/home/data/data1.txt' into table tab1;

1.问题：（top10）按照来源排序，访问量高的排最上面

select source,count(*) num
from tab1
group by source
order by num desc;

select 查询在order by 前

 

2.问题：（推荐系统）给一个用户ip地址，得到用户经常访问的应用类型后，推荐用户同种类型的其他应用      

数据二：

建表，填数据：

create table if not exists tab2(
  id      string,
  name    string,
  url     string,
  pid     string
)
row format delimited fields terminated by '\t'
stored as textfile;

load data local inpath '/home/data/data2.txt' into table tab2;

1.问题：(数据清洗)合并name与url，格式为 NAME:name|URL:url

select concat('NAME:',name,'|','URL:',url)
from tab2
where name is not null and  url is not null;

数据三：

表的建立和数据插入

create table if not exists tab4(
  no        string,
  province  string,
  city      string,
  pid	    string,
  cid	    string
)
row format delimited fields terminated by '\t'
stored as textfile;

load data local inpath '/home/data/data4.txt' into table tab4;

1.问题：从源数据中筛出pid与省份、cid与城市，并且创建新表保存 去重 distinct 

这里使用加行键的方法，实行唯一标识。

select pid,province
from tab4
group by pid,province
limit 1

select t1.cid,t1.city,t1.rank
from(
select cid,city,row_number() over (partition by cid order city) rank
from tab4
group by cid,city) t1
where t1.rank=1;

数据四：

新建表以及添加数据

create table if not exists tab5(
  lac	 string,
  cellid	 string,
  cell_name	 string,
  longitude	 string,
  latitude	 string
)
row format delimited fields terminated by '\t'
stored as textfile;

load data local inpath '/home/data/data5.txt' into table tab5;

1.问题：从cell_name列中截取需要的部分，例如 HZD0090，截取D0090杭州四季青蔬菜公司，截取D0090

    

数据五：

建表+添加数据

create table if not exists tab6(
  id	 string,
  service	 string
)
row format delimited fields terminated by '\t'
stored as textfile;

load data local inpath '/home/data/data6.txt' into table tab6;

问题：去除所有父类服务，只要子类服务（id 是字符串类型）

select * from tab6 where id>100

数据六：

添加数据：

create table if not exists tab7(
  id	 string,
  type	 string,
  sagem	 string
)

row format delimited fields terminated by '\t'
stored as textfile;

load data local inpath '/home/data/data7.txt' into table tab7;

问题：按照设备类型，统计出现的频率

select type,count(*)
from tab7
group by type;

数据七：

问题：去重后存入到新表中

select col,row_number() over (partition by col order by col) rank
from tab10
group by col;