当前位置：首页 > article >正文

hive-sql 计算每年在校生人数

article 2025/1/4 16:23:42

sql：

用到函数 space() ：生产空格 split 切分空格

posexplode 炸裂函数返回下标和具体指

with data as (
--             主键        入学年份            人数            学制
    select 1 as id , 2001 as yy , 1200 as stu_num , '3' as stu_len union all
    select 2 as id , 2000 as yy , 1300 as stu_num , '5' as stu_len union all
    select 3 as id , 2003 as yy , 1400 as stu_num , '4' as stu_len union all
    select 4 as id , 2002 as yy , 1500 as stu_num , '3' as stu_len union all
    select 5 as id , 2002 as yy , 1600 as stu_num , '4' as stu_len union all
    select 6 as id , 2014 as yy , 1600 as stu_num , '4' as stu_len union all
    select 7 as id , 2015 as yy , 1600 as stu_num , '4' as stu_len union all
    select 8 as id , 2015 as yy , 1600 as stu_num , '4' as stu_len union all
    select 9 as id , 2015 as yy , 1600 as stu_num , '4' as stu_len union all
    select 10 as id , 2015 as yy , 1600 as stu_num , '4' as stu_len union all
    select 11 as id , 2016 as yy , 1600 as stu_num , '4' as stu_len union all
    select 12 as id , 2016 as yy , 1600 as stu_num , '4' as stu_len union all
    select 13 as id , 2016 as yy , 1600 as stu_num , '4' as stu_len union all
    select 14 as id , 2016 as yy , 1600 as stu_num , '4' as stu_len union all
    select 15 as id , 2016 as yy , 1600 as stu_num , '4' as stu_len 
)
SELECT
  endyy,
  SUM(stu_num)
FROM
  (
    SELECT
      id,
      (yy + pos) AS endyy,
      stu_num,
      stu_len,
      val
    FROM
      data LATERAL view posexplode(split(space(CAST(stu_len AS INT) -1),' ')) t AS pos,
      val
  ) t
GROUP BY
  endyy

结果：