oracle根据多个字段删除重复记录

根据GOODS_NUMBER,USER_ID分组, 然后组内根据CREATED_DATE排序,生成每条记录的排序号rn,取每组中rn>1的记录为重复记录, 最后根据id删除

delete  HNZSX_ORDER WHERE id in  
(
  select id from (
         select t.*, row_number() over(partition by GOODS_NUMBER,USER_ID order by CREATED_DATE desc) rn from HNZSX_ORDER t 
  ) where rn > 1
)