hive內group by取第一條數據，Hive中row_number的使用

一、hive的分組和組內排序---語法

語法：
row_number() over (partition by 字段a order by 計算項b desc ) rank

• rank是排序的別名
• partition by：相似hive的建表，分區的意思；
• order by ：排序，默認是升序，加desc降序；
• 這裏按字段a分區，對計算項b進行降序排序

二、hive的分組和組內排序 --- 實例

要取top10品牌，各品牌的top10渠道，各品牌的top10渠道中各渠道的top10檔期

一、取top10品牌

select 「品牌」 , sum/count/其餘() as num from "table_name" order by num desc limit 10;

二、取top10品牌下各品牌的top10渠道

select a.* from (select "品牌","渠道",sum/count() as num, row_number () over (partition by "品牌" order by num desc) rank from 「table_name」 where 品牌限制條件 group by 「品牌」,「渠道」 ) a having a.rank <= 10;

三、 取top10品牌下各品牌的top10渠道中各渠道的top10檔期

select a.* from (select "品牌","渠道","檔期",sum/count/其餘() as num row_number() over (partition by "檔期" order by num desc) rank from "table_name" where 品牌限制條件 group by 「品牌」,「渠道) a Having a.rank <= 10;

個人應用：rank編號，rank小組內從1開始編號

SELECT v.visitor_phone,v.city，v.bigarea, row_number()over (partition by visitor_phone order by visitor_phone desc) rank from visitor v WHERE 1=1 and visitor_name in('蔣鳳','周金魁') group BY v.visitor_phone,v.city
---------------------------------
18222666666 北京 華北區 1
18222666666 天津 華北區 2
13402777777 北京 華北區 1
13402777777 成都 中西部 2

取編號爲1的

SELECT a.* from( SELECT v.visitor_phone,v.city，v.bigarea, row_number()over (partition by visitor_phone order by visitor_phone desc) rank from visitor v WHERE 1=1 and visitor_name in('蔣鳳','周金魁') group BY v.visitor_phone,v.city )a where a.rank=1;-------------------------------------18222666666 北京 華北區   113402777777 北京 華北區	1