Re: [boost] RTL

From: Jose (jmalv04_at_[hidden])
Date: 2005-10-20 07:31:18

Hi,

Find attached the results for RTL.

Calum, I need your output to provide the a 3rd column with the AS name (like
RTL output does) so that the query results are identical.
There seems to be a mismatch in the number of records: RTL shows 21598
records loaded and RML 19439.

Also, the update queries need to be the same. The update of the query
consists of deleting the first record in the result: AS 3317
The number of records deleted in the update for RTL (1982) is irrelevant.
The two benchmark times that matter is the query time and the update time.

For RTL: 11 sec and 330 ms

Regards
Jose

Regards
On 10/18/05, Arkadiy Vertleyb <vertleyb_at_[hidden]> wrote:
>
> "Calum Grant" <calum_at_[hidden]> wrote
>
> > > This sounds like a correct result -- can I see the first 50 items?
> >
> > The output is at the bottom.
>
> This result is not sorted by (counter, city) as requested, which makes it
> impossible to compare it with mine.
>
> > I've managed to get my times down a bit by reimplementing the solution
> > in a much more straightforward way. I can insert all 21421 items,
> > calculate the number of neighbours, and display the top 500 in 235ms.
> > Then I can remove 1982 items,
>
> from where?
>
> > and redisplay the results, in 47
> > milliseconds. This of course makes me suspicious that I'm doing
> > something wrong.........
>
> This numbers are very impressive, but they don't tell much unless Jose
> runs
> it on his setup (after the result is properly sorted).
>
> > The solution I have is not terribly elegant from RML's point of view,
> > since as I said before RML is designed for logical queries rather than
> > numerical computation. It was not a terribly interesting problem
> > because it needed just a single table with two indexes.
>
> Hmm, I managed to program it in RTL as a single relational view on the
> original table, without doing _anything_ by hand. The correctness of this
> view is preseved by the library when the data is changed. In your case
> _you_ remove 1982 items. In my case _the library_ removes 1982 items. I
> think your solution is rather by-hand than RML-based.
>
> Can you create a single RML query on the original table that solves the
> problem? Can you claim that using RML added a lot to your solution? Is it
> faster/more elegant/readable/etc. than just using STL directly?
>
> Regards,
> Arkadiy
>
>
>
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost
>


   Hi,
   Find attached the results for RTL.
   Calum, I need your output to provide the a 3rd column with the AS name
   (like RTL output does) so that the query results are identical.
   There seems to be a mismatch in the number of records: RTL shows 21598
   records loaded and RML 19439.
   Also,  the update queries need to be the same. The update of the query
   consists of deleting the first record in the result
   Also,  the  number of records deleted as a result of the update should
   not be hand-coded in the program
   Regards

   On 10/18/05, Arkadiy Vertleyb <[1]vertleyb_at_[hidden]> wrote:

     "Calum Grant" <[2]calum_at_[hidden]> wrote
     >  >  This  sounds  like a correct result -- can I see the first 50
     items?
     >
     > The output is at the bottom.
     This  result  is  not sorted by (counter, city) as requested, which
     makes it
     impossible to compare it with mine.
     >  I've  managed  to  get my times down a bit by reimplementing the
     solution
     >  in  a  much  more  straightforward  way.  I can insert all 21421
     items,
     >  calculate  the  number of neighbours, and display the top 500 in
     235ms.
     > Then I can remove 1982 items,
     from where?
     > and redisplay the results, in 47
     > milliseconds.  This of course makes me suspicious that I'm doing
     > something wrong.........
     This  numbers  are very impressive, but they don't tell much unless
     Jose runs
     it on his setup (after the result is properly sorted).
     >  The  solution I have is not terribly elegant from RML's point of
     view,
     > since as I said before RML is designed for logical queries rather
     than
     > numerical computation.  It was not a terribly interesting problem
     > because it needed just a single table with two indexes.
     Hmm,  I managed to program it in RTL as a single relational view on
     the
     original  table, without doing _anything_ by hand.  The correctness
     of this
     view  is preseved by the library when the data is changed.  In your
     case
     _you_  remove  1982  items.  In  my case _the library_ removes 1982
     items.  I
     think your solution is rather by-hand than RML-based.
     Can you create a single RML query on the original table that solves
     the
     problem?  Can  you  claim  that  using  RML  added  a  lot  to your
     solution?  Is it
     faster/more elegant/readable/etc.  than just using STL directly?
     Regards,
     Arkadiy
     _______________________________________________
     Unsubscribe & other changes:
     [3]http://lists.boost.org/mailman/listinfo.cgi/boost

References

   1. mailto:vertleyb_at_[hidden]
   2. mailto:calum_at_[hidden]
   3. http://lists.boost.org/mailman/listinfo.cgi/boost