Initialization

The class is initialized by passing the following arguments:

regression_procedure

- a fully qualified name of the module/class in which the regression is implemented (see Section 6.5.5). Default value is ``opus_core.linear_regression''.

filter

- name of a variable/attribute used to filter out elements for the regression (applied to both, the run() and estimate() method). Default value is ``urbansim.gridcell.is_in_development_type_group_developable'' which requires an entry 'developable' in the table ``develpment_type_groups'' ().

submodel_string

- If model contains submodels, this character string specifies what attribute of the observation set determines those submodels. Default value is ``development_type_id''.

run_config

- A collection of additional arguments that control a simulation run. It should be of class Resources.

estimate_config

- A collection of additional arguments that control an estimation run. It should be of class Resources.

The constructor sets filter as a class attribute and calls the parent constructor (see Initialization in Section 6.4.4).