[Mondrian] mondrian.rolap.aggregates.ChooseByVolume Property inMondrian

John V. Sichi jsichi at gmail.com
Wed Apr 18 11:30:18 EDT 2007


Or...

http://code.google.com/soc/

JVS

Richard Emberson wrote:
> This might make a great undergraduate CS senior project:
> Mondrian volume vs. nos or rows, vs. database (mysql, oracle, etc.),
> vs. size of aggregate table, etc.
> The student would come out with niche knowledge - which
> can help in getting a meaningful first job.
> Anyone know any CS undergraduates or their advisors?
> 
> Julian, if Mondrian had a web page listing, say, 5 to 10
> investigations that college students could do with
> Mondrian, it might sow the seeds of future contributors.
> 
> Just a very early, pre-coffee, morning thought.
> 
> 
> Richard
> 
> 
> Julian Hyde wrote:
>> No one ever studied it, to my knowledge. This is one of those cases 
>> where the developer (in this case Richard Emberson) gives the end-user 
>> a choice because he doesn't know the answer.
>>  
>> My gut would be to go with row count, because it is simpler.
>>  
>> Julian
>>
>>     
>> ------------------------------------------------------------------------
>>     *From:* mondrian-bounces at pentaho.org
>>     [mailto:mondrian-bounces at pentaho.org] *On Behalf Of *JIE WU
>>     *Sent:* Tuesday, April 17, 2007 9:13 PM
>>     *To:* mondrian at pentaho.org
>>     *Subject:* [Mondrian] mondrian.rolap.aggregates.ChooseByVolume
>>     Property inMondrian
>>
>>     Hi,
>>
>>     
>>     As suggested in the Mondrian spec,
>>     /mondrian.rolap.aggregates.ChooseByVolume/ determines whether to
>>     choose aggregate tables based on volumn or rows. My question is that
>>     *is there any study on this property to tell which method (ByVolumn
>>     or ByRows) to choose aggTables is better under which circumstances*?
>>     If so, has Mondrian been evaluated against this property? What is
>>     the result like? This information would be quite useful when coming
>>     to tuning Mondrian for selecting aggregate tables?
>>
>>     
>>     Can anybody shed some light on this? I would be greatly appreciative
>>     if you could help.
>>
>>     
>>     Thanks
>>
>>     Jie
>>
>>     
>>     ===================
>>
>>     Jie Wu
>>
>>     FIT 410, The University Of Memphis
>>
>>     Memphis, TN, 38111, USA
>>
>>     www.jiewu.info
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Mondrian mailing list
>> Mondrian at pentaho.org
>> http://lists.pentaho.org/mailman/listinfo/mondrian
> 
> 




More information about the Mondrian mailing list