Hedonic house price models typically impose a constant price structure on housing factors throughout an entire market area. However, there is increasing evidence that the prices of many important factors vary over space. The Geographically Weighted Regression (GWR) of the spatial non-stationary model is used to explore the factors that impact on the housing price in Chongqing urban areas. The result was compared with OLS model. It demonstrated that GWR model was better than the OLS mod-el. The relationships of housing price and factors changed with space position positively and negatively. Therefore analyzing the causes of housing price according to the local conditions is necessary for optimizing the spatial structure of housing price and guid-ing the orderly expansion of city.% 针对普通最小二乘法(OLS)从空间全局角度分析的不足,运用地理加权回归模型(GWR)探索重庆市主城区各影响因素在不同空间位置对房价的作用机理。结果表明, GWR模型显著优于OLS模型,是定量研究各因素在不同空间位置对房价经济贡献价值的有效方法;交通设施、商业集聚中心、公共服务设施、自然环境是影响房价的重要因素,各影响因素与房价的关系随空间位置而发生正负向变化,影响程度有显著的空间差异,说明只有因地制宜的分析房价成因,才能优化城市房价的空间结构,引导城市有序扩展。
展开▼