Real Estate Price Model (REPM)

Objective

The Real Estate Price Model (REPM) predicts the price per unit of each building.

Algorithm

It uses a hedonic regression structure, which is a multiple regression, estimated using Ordinary Least Squares (OLS), normally with the price specified as a log of price.

Configuration

OPUS path to model code

urbansim.models.real_estate_price_model

Model Parameters

The configuration of the REPM in the zone model system is summarized in the following table:

Element	Setting
Dataset	Buildings
Dependent Variable	Log of Price Per Unit (per housing unit for residential, per square foot for non-residential buildings)
Model Type	Regression
Submodels	Building Type - separate models are specified for each type of building
Independent Variables	Constant, and attributes of building: density, accessibility, zonal composition of households and employment

Data

These tables are used by the Real Estate Price Model in the zone-based version of UrbanSim.

Table Name	Brief Description
buildings	Aggregated buildings, by building_type and zone
zones	Zones used in the travel model, for accessibility and density variables
travel_data	Zone-to-zone skims from the travel model, for accessibility variables
households	Household data, for socioeconomic and density variables
jobs	Employment data, for accessibility and density variables

-- PaulWaddell - 07 Dec 2009