spark操作map数据

最新推荐文章于 2024-06-02 10:08:44 发布

盛源_01

最新推荐文章于 2024-06-02 10:08:44 发布

阅读量135

点赞数

文章标签： spark 大数据分布式

本文链接：https://blog.csdn.net/weixin_40829577/article/details/133274271

版权

一转换到map操作

1 行数据聚合成map

with init_data as ( -- 构造数据源
    select '11' as id, 'app_1' as pkg, 1 as is_init
    union all
    select '11' as id, 'app_2' as pkg, 0 as is_init
    union all
    select '22' as id, 'app_3' as pkg, 0 as is_init
    union all
    select '22' as id, 'app_4' as pkg, 0 as is_init
)
select
    id,
    map_from_entries( collect_list(struct(pkg, is_init)) ) as info
from
    init_data
group by
    id
;