start_run
script:
gridcells tables by unrolling the
develoment_event_history data.
The unrolling processes all records in the development_event_history
table. It starts with the most current year of records, and moves backward in
time. For this description, assume that the baseyear is 2000. The unrolling
process first loads the gridcell dataset for year 2000. For each
develoment_event_history record with scheduled_year for the prior
year (e.g. for 1999) it substracts from the gridcells any development by that
event (e.g. residential_units). Values that go negative are set to
zero. Once all events for this year have been ``undone'', the unrolling
process writes the modified gridcell dataset to the prior year, e.g. to 1999.
This process is repreated for each year with data in
develoment_event_history.
If you only want 10 years of data to be unrolled, only put those years of data
into your develoment_event_history table.
run_activity table in the services
database, using a new run_id value unique to this run. In order to help
match runs with their cache directory, the name of the cache directory begins
with the run_id value, e.g. run_342.2006_04_25_09_40.
'models' section of the
'travel_model_configuration'. Typically, it includes a preparation step
that prepares travel model inputs from the UrbanSim data, a step that
actually runs the travel model, and one or more steps to puts the results of
the travel model into the travel_data dataset for the current year, so
the data is visible to the UrbanSim models in that and following years. Each
model step is done in a separate process. The log for the running of the
model is written to a file named, e.g., emme2_2005_log.txt.
run_activity table, a row is
either added or updated in the available_runs table in the
services database. This table has a single row per run and records the
row's current state and information.