[Mondrian] Aggregate tables and distinct-count

Luc Boudreau lucboudreau at gmail.com
Mon Jan 14 12:14:47 EST 2013


Be careful with the non-collapsed dimensions. This feature needs the tables
to be in a snowflake form. If the dimension table is a single table and you
join on the non-last level, you will get a cartesian product.

Luc

On Mon, Jan 14, 2013 at 12:04 PM, Pedro Alves <pmgalves at gmail.com> wrote:

>
>
> IT IS WORKING!
>
>
> However, I had to skip the first level of a hierarchy, marking the
> second level as collapsed= false.. something I've never used before and
> seems to be exactly for this :p
>
>
>
> -pedro
>
>
>
>
> On 01/14/2013 05:01 PM, Julian Hyde wrote:
> > yes, it should be used.
> >
> > furthermore, mondrian knows that it is safe to roll up any attributes
> that are functionally dependent on the measure column. (e.g. if you have
> computed count(distinct customer_id), and you have in the agg table the
> distinct count for male and female customers in each state, you can compute
> the number of distinct customers in each state by adding the distinct males
> and females.)
> >
> > Julian
> >
> >
> > On Jan 14, 2013, at 8:52 AM, Pedro Alves <pmgalves at gmail.com> wrote:
> >
> >>
> >>
> >> I know we can't aggregate distinct-count measures like we do with the
> >> others.
> >>
> >> But if a query matches a specific aggregation table's bitkey,
> >> couldn't/shuoldn't it be used?
> >>
> >>
> >>
> >> -pedro
> >> _______________________________________________
> >> Mondrian mailing list
> >> Mondrian at pentaho.org
> >> http://lists.pentaho.org/mailman/listinfo/mondrian
> >
> > _______________________________________________
> > Mondrian mailing list
> > Mondrian at pentaho.org
> > http://lists.pentaho.org/mailman/listinfo/mondrian
> >
> _______________________________________________
> Mondrian mailing list
> Mondrian at pentaho.org
> http://lists.pentaho.org/mailman/listinfo/mondrian
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.pentaho.org/pipermail/mondrian/attachments/20130114/a183b8f8/attachment.html 


More information about the Mondrian mailing list