If the measure is not pre-computed, the query sent to Kylin would get
exception, not reroute to hive.

2016-08-17 7:52 GMT+08:00 Sandeep Khurana <[email protected]>:

> It will go to hive and execute the query, right? Since this measure is not
> pre computed
>
> On 17-Aug-2016 4:37 am, "Li Yang" <[email protected]> wrote:
>
>> Not sure I understand the problem. Since dimension table CITY is merged
>> into fact, there must be a CITY column on fact table? Then why not
>> SUM(SPENDING) GROUP BY CITY?
>>
>> On Tue, Aug 9, 2016 at 1:40 AM, Sandeep Khurana <[email protected]>
>> wrote:
>>
>>> Hello
>>>
>>> I have a big dimension table which I merged into fact table.  If it were
>>> not merged then I could create measures on the join key. Now since it is
>>> merged with fact table, how can  I still create the measure within the same
>>> merged fact ?
>>>
>>> e.g.  if sales is fact table and there are 2 dimension tables customer
>>> and city. I merge customer to sales table as customer table has very large
>>> number of rows. After merging how can I execute query like 'city wise
>>> spending (sum) of the customer' . How will I define measures for this on
>>> merged fact?
>>>
>>>
>>


-- 
With Warm regards

Yiming Liu (刘一鸣)

Reply via email to