当我使用分组集（多维数据集）时，为什么不能通过stmt来过滤grouping__id？

Question

hive code as follows:

set mapred.reduce.tasks = 100;
create table order_dimensions_cube as
select
        grouping__id as groupid,
        user_level             ,
        city_level             ,
        region_name            ,
        province_name          ,
        city_name              ,
        platform               ,
        sale_type              ,
        item_first_cate_name   ,
        app_module             ,
        department             ,
        sum(COALESCE(complete_sum, 0)) as complete_price
from
        data
group by
        user_level          ,
        city_level          ,
        region_name         ,
        province_name       ,
        city_name           ,
        platform            ,
        sale_type           ,
        item_first_cate_name,
        app_module          ,
        department
with cube having grouping__id >= 704;

this turns out that no records generated.

more info:

1. I checked that I have a lot of records in table:data.
2. I have tried this sql without the having stmt and there is alot records generated.

why this happens and how to solve this if I want to use having to do some constraints on the result?

thank you.

Answer 1

since you did not provide actual data, please try the following:

select grouping_id,count(*) from 
(select
        grouping__id as groupid,
        user_level             ,
        city_level             ,
        region_name            ,
        province_name          ,
        city_name              ,
        platform               ,
        sale_type              ,
        item_first_cate_name   ,
        app_module             ,
        department             ,
        sum(COALESCE(complete_sum, 0)) as complete_price
from
        data
group by
        user_level          ,
        city_level          ,
        region_name         ,
        province_name       ,
        city_name           ,
        platform            ,
        sale_type           ,
        item_first_cate_name,
        app_module          ,
        department 
with cube) A
group by grouping_id

and see how many records are there for each grouping__id. there could be some issues there. also - try changing the outer query to

select * from 
(select
        grouping__id as groupid,
        user_level             ,
        city_level             ,
        region_name            ,
        province_name          ,
        city_name              ,
        platform               ,
        sale_type              ,
        item_first_cate_name   ,
        app_module             ,
        department             ,
        sum(COALESCE(complete_sum, 0)) as complete_price
from
        data
group by
        user_level          ,
        city_level          ,
        region_name         ,
        province_name       ,
        city_name           ,
        platform            ,
        sale_type           ,
        item_first_cate_name,
        app_module          ,
        department 
with cube) A
where grouping__id >= 704 

and see if problem persists..

this is not a solution but more of a trial to understand what goes