Hana,
We send you our "old" information as well as something new.
Thank you for your reply, but we checked all the tables. There are no grid_id values in the household table and in any other table that do not exist in the gridcell table. The same was checked for zone_id values. So, it should be other reason of the error.
We wrote the following script to check if simulation is sensible to zone_id values. In the Eugene database, we increased zone_id values where they were more than 100:
UPDATE Zones Z
Set Z.zone_id = Z.zone_id+650000000 where Z.zone_id > 100 ;
Commit;
UPDATE gridcells Z
Set Z.zone_id = Z.zone_id+650000000 where Z.zone_id > 100 ;
Commit;
UPDATE travel_data Z
Set Z.from_zone_id = Z.from_zone_id+650000000 where Z.from_zone_id > 100 ;
Commit;
UPDATE travel_data Z
Set Z.to_zone_id = Z.to_zone_id+650000000 where Z.to_zone_id > 100 ;
Commit;
While running simulation, we obtained the same errors as described before. So, zone_id values seem to be important. We suppose that zone_id values are used for access 'numarray" data (as commented in the method aggregate_over_ids() of opus_core/dataset.py).
Hakim,
We checked again our tables. Grid_id values are not duplicated in the gridcells table and in other tables. Zone_id values are not duplicated in any table as well.
Thanks for your responce,
Philippe and Mark
> --- Èñõîäíîå ñîîáùåíèå ---
> Îò êîãî: Hana Sevcikova (hana_at_cs.washington.edu)
> Êîìó: Mark Kryvobokov
> Äàòà: 27 ñåíòÿáðÿ, 02:58:07
> Òåìà: Re: [UrbanSim-Users] Own identifiers: zone_id, grid_id, group_id,etc.
>
> > Mark,
> >
> > I can handle the first part of your question:
> >
> > The error comes due to a mismatch between grid_id in gridcells and
> > grid_id in households. I assume there are some values in the households
> > table that do not exist in the gridcells table. Please check.
> >
> > It should not be connected to the fact that there is a big difference
> > between the max and min of your zone_id. However, the code will run more
> > efficiently, if the difference is small (about twice of the number of
> > elements in the table). This fact applies to any dataset.
> >
> > Hana
> >
> >
> >
> > On Tue, 25 Sep 2007, Mark Kryvobokov wrote:
> >
> > > Hello!
> > >
> > > We still use UrbanSim version 4.0 (though we are thinking about the updating).
> > >
> > > We have a question about the creation of our own identifiers. We tried to use our own zone_id (unique integer values with the range of 10050000 to 693890604) . We changed the Eugene zone_id values with our zone_id values in zones and gridcells tables. We changed also from_zone_id and to_zone_id values in travel_data table. After that simulation running failed while executing the Commercial Employment Location Choice Model. This is a part of the traceback message:
> > >
> > >
> > > File "C:\MyNew\projects\OPUS\opus-4.0\urbansim\gridcell\population.py", line 34, in compute
> > >
> > > return self.get_dataset().sum_dataset_over_ids(households, self.hh_persons)
> > >
> > > File "C:\MyNew\projects\OPUS\opus-4.0\opus_core\dataset.py", line 1118, in sum_dataset_over_ids
> > >
> > > return self.aggregate_dataset_over_ids(dataset, "sum", attribute_name, constant)
> > >
> > > File "C:\MyNew\projects\OPUS\opus-4.0\opus_core\dataset.py", line 1148, in aggregate_dataset_over_ids
> > >
> > > return self.aggregate_over_ids(dataset_id_values[used], values, function)
> > >
> > > File "C:\MyNew\projects\OPUS\opus-4.0\opus_core\dataset.py", line 1112, in aggregate_over_ids
> > >
> > > raise StandardError, "Unknown function " + function + " or error occured during evaluation."
> > >
> > > StandardError: Unknown function sum or error occured during evaluation.
> > > We suppose that the reason or the error is a big difference between max and min values of zone_id. Could somebody confirm this and/or explain why it is so?
> > >
> > > Our next question is about group_id. In development_type table for Eugene, we see min_sqft and max_sqft for different development types. Is there some rule according to which a gridcell should belong to e.g. industrial or commercial group if there are different types of development? Can we formulate our own rules?
> > >
> > > Thank you in advance,
> > > Mark and Philippe
> >
>
>
> --
> ÂÀØÅ 100% ÇÐÅÍÈÅ. Îáðóäîâàíèå ¹1 â Óêðàèíå. Óçíàé ïîäðîáíåå.
> http://www.ailas.com.ua
>
>
> _______________________________________________
> Users mailing list
> Users_at_urbansim.org
> http://www.urbansim.org/mailman/listinfo/users
>
>
--
Ïðèõîäèòå íà áåñïëàòíûé ñåìèíàð, ðàññêàæåì êàê çàðàáàòûâàòü íà âàëþòíîé áèðæå
Ïåðåõîä http://www.fxclub.org/filials_kiev_reg
Received on Mon Oct 01 2007 - 06:14:49 PDT
This archive was generated by hypermail 2.2.0 : Mon Oct 01 2007 - 06:14:51 PDT