hive,spark获取TOPN窗口函数

TOPN row number 说明：  row_number() over ([partition col1] [order by col2])  rank() over ([partition col1] [order by col2])  dense_rank() over ([partition col1] [order by col2])  它们都是根据col1字段分组，而后对col2字

>>阅读原文<<