The Run Method

Input:

model_configuration

- a Configuration object that determines project types and settings for each type (see Section Model Configuration bellow).

vacancy_table

- a Dataset object that must have attribute ``year'' as the unique identifier and should contain attributes ``target_total_residential_vacancy'' and ``target_total_non_residential_vacancy''. Those attributes give vacancy rates for specific years.

history_table

- a Dataset object giving historical data of development events.

year

- simulated year.

location_set

- a Dataset object on which vacancies will be determined.

resources

- a Resources object for additional arguments. It can contain names of variables that compute the current vacancies per project type. Those entry names are composed from the project type and a character string `vacant_variable', e.g. for commercial project type the key for this argument is ``commercial_vacant_variable''. Default values for those entries are ``urbansim.gridcell.vacant_'' + project type, e.g. ``urbansim.gridcell.vacant_commercial''.

Algorithm: 
The method first determines the target vacancy rates. It distinguishes between residential and non-residential vacancy rate. For each project type given in the model_configuration, the method obtains current vacancy in each location (from location_set and the overall vacancy rate). By comparing the rate with the target vacancy rate, it determines if there is a need for development and how many units should be developed. If there are units to be developed, from the historical data it samples sizes of projects so that their sum gives approximately the desired number of units to be developed. It then creates an instance of class DevelopmentProjectDataset for this particular project type that has unique identifier called ``project_id''. Each project corresponds to one of the sampled amount of units. This amount is stored in the new dataset as an attribute given in the entry ``units'' of model_configuration (see below for details). If an attribute called project type + ``_improvement_value'' is contained in the location_set, an average improvement value is computed over all locations and this project type. This value is then added to each project in the created DevelopmentProjectDataset. Additionally, an attribute of the same name as the unique identifier of location_set is added to the DevelopmentProjectDataset and set to 0's (which means that the projects are unplaced).

Output: 
The method returns a dictionary with an entry for each project type containing the corresponding DevelopmentProjectDataset object. If no projects were created for a project type, the corresponding dictionary value is None.