Re: [UrbanSim-Users] ELCM and scalable jobs

Maybe you're refering to the fact that some jobs have grid_id=0 (without
actually having such gridcell in the gridcells table)? In that case it
means that those jobs are unplaced.

Hana

Hana Sevcikova wrote:
> Mark,
>
> I don't see anything in the log file why there should be a new gridcell.
> In fact, no model of urbansim produces new gridcells. Could you find out
> exactly after which model the gridcell appears?
>
> Thanks,
>
> Hana
>
> Mark Kryvobokov wrote:
>>
>>
>> Thank you Hana,
>>
>> I send my log file for the first simulation year.
>>
>> Mark
>>
>> --- Original Message ---
>> From: Hana Sevcikova <hana_at_cs.washington.edu>
>> To: Mark Kryvobokov <k_mark_at_ukr.net>
>> Date: 4 march, 22:33:30
>> Subject: Re: [UrbanSim-Users] ELCM and scalable jobs
>>
>> Mark,
>>
>> > In our
>> > run_multiprocess.log file, I don't see anything about capacity.
>> Is it
>> > the log file, which you mean?
>>
>> I mean the log files for each year (e.g. year_2010_log.txt). In the
>> ELCM
>> sections you would see how many jobs were not placed. You would see
>> something like
>>
>> Simulate ...
>> Number of unplaced agents: 150 (in 3 iterations)
>>
>> But your case is the one with missing specification for some
>> sub-models.
>>
>> >
>> > Where is a specification of the Scaling Jobs Model? In the
>> > scaling_jobs_model.py, I don't see the description of variables.
>>
>> This model does not have any specification. It distributes jobs
>> according to the current distribution of sectors.
>> (See http://www.urbansim.org/docs/opus-userguide/node217.html)
>> It considers all sectors that appear among the governmental jobs.
>>
>> >
>> > One new question. After the first year simulation, a new gridcell
>> with
>> > grid_id=0 appears. It includes about 13% of households and
>> population.
>> > What is it? It did not exist in the baseyear.
>>
>> I don't know what it is, I haven't heard about such case before.
>> Please
>> send the log file for the first year.
>>
>> > And one more question related to your answers to others: I
>> don't see
>> > your comments in
>> > household_location_choice_model_configuration_creator.py and
>> > filter_for_estimation in
>> land_price_model_configuration_creator.py. I'm
>> > looking at files in
>> >
>> >
>>
>> https://trondheim.cs.washington.edu/svn/opus/tags/4.1.2/urbansim/configurations/household_location_choice_model_configuration_creator.py
>>
>>
>> >
>> >
>> > Or they are in another place?
>>
>> It's at
>>
>> https://trondheim.cs.washington.edu/svn/opus/trunk/urbansim/configurations/...
>>
>>
>>
>> Hana
>>
>> >
>> > --- Original Message ---
>> > From: Hana Sevcikova <hana_at_cs.washington.edu>
>> > To: Mark Kryvobokov <k_mark_at_ukr.net>
>> > Date: 4 march, 01:45:16
>> > Subject: Re: [UrbanSim-Users] ELCM and scalable jobs
>> >
>> > Mark,
>> >
>> > The jobs are separated into industrial, commercial, home-based and
>> > scaling (i.e. governmental) type according to their building
>> type. In
>> > your case:
>> >
>> > 152317 + 152442 + 73647 + 151697 = 530103
>> >
>> > This corresponds to the number of movers that the relocation model
>> > determines.
>> >
>> > Within these four models the jobs are processed in sub-models
>> which are
>> > determined by their sector ids. (The Scaling Job Model does not
>> have
>> > sub-models but is also processed by sectors.)
>> >
>> > The Distribute Unplaced Jobs Model then takes all jobs that are
>> still
>> > unplaced after running those four models and places them
>> according to
>> > the current distribution of jobs by sector.
>> >
>> > If you get 228799 unplaced jobs it can be for two reasons:
>> >
>> > 1. There is not enough capacity (you should be able to find out
>> from
>> > the
>> > log file)
>> >
>> > 2. Your specification does not include all sub-models. For
>> example, if
>> > your specification for commercial ELCM contains sub-models 1,2,
>> and 3,
>> > but there are commercial jobs of sectors 4 and 5, those jobs
>> would not
>> > be considered by any of the ELCMs, neither by the scaling jobs
>> model -
>> > they would be located by the Distribute Unplaced Jobs Model. But
>> they
>> > would be included in the sum of the 'total number of
>> individuals' in
>> > the
>> > commercial ELCM.
>> >
>> > Does it help?
>> >
>> > Hana
>> >
>> > Mark Kryvobokov wrote:
>> > >
>> > >
>> > > Hello!
>> > >
>> > > Our simulation is running without errors. I want to understand
>> some
>> > > points there:
>> > >
>> > > 1. What are 'total number of individuals' in industrial,
>> > commercial and
>> > > home-based ELCM? We have the following total numbers: 152317
>> > industrial,
>> > > 152442 commercial and 73647 home-based. The sum of them all is
>> > 378406,
>> > > which is much lower than the number of movers that equals to
>> > 530103 (the
>> > > result of the Employment Relocation Model). As I understand, we
>> have
>> > > 530103 jobs (new and relocated), which should be moved to
>> > gridcells. The
>> > > 'total number of individuals' in ELCM is not the same?
>> Industrial,
>> > > commercial and home-based jobs are distinguished by
>> > building_type? In
>> > > this case governmental building_type is not considered?
>> > >
>> > > 2. What is 'number of scalable jobs' in the Scaling Jobs Model
>> > and the
>> > > Distribute Unplaced Jobs Model? We have 151697 and 228799
>> > respectively.
>> > > As I understand, jobs are relocated within sectors according to
>> > > distribution of existing jobs in the former model and
>> randomly in
>> > the
>> > > latter model. At the same time I see that the building_type is
>> > used in
>> > > both cases. So, jobs are relocated considering sector_id or
>> > > building_type or both? In our case I again can't find a link
>> > between the
>> > > number of jobs, which should be moved (within sectors or
>> building
>> > > types), and the numbers of scalable jobs.
>> > >
>> > > Mark
>> > >
>> > >
>> > >
>> >
>>
>> ------------------------------------------------------------------------
>>
>> >
>> > >
>> > > _______________________________________________
>> > > Users mailing list
>> > > Users_at_urbansim.org
>> > > http://www.urbansim.org/mailman/listinfo/users
>> >
>>
>
Received on Fri Mar 07 2008 - 18:48:31 PST