The House Price Spatial Distribution Pattern Analysis Based on Cokriging:the Case of Changsha
以长沙为例，比较克里格法和协同克里格法在房价空间分布插值中的精度，发现协同克里格法更准确，房价呈同心椭圆分布，南北方向扩散明显。
The regional differences of house price distribution and the variance degree indicate that the spatial stability condition of house price regression analysis is not satisfied, so the consideration of regional variables in discussing the house price spatial distribution difference has become a new direction of studying house price. In order to examine the house price distribution pattern from region's multi-attributes, and enhance the evaluation accuracy, this paper, based on spatial auto correlation analysis, takes Changsha city as the case to comparatively study its house price spatial distribution pattern by interpolated maps from Kriging and Cokriging methods. The results show: 1) Cokriging method, with the consideration of regional multi-attributes, is more accurate in interpolating prediction than that of Kriging; 2) The house price of Changsha city is significant in spatial correlation, and the directions of south- north and east- west are secondorder decreasing from core to periphery, so the spatial variance of prices are heterotopic in their directions; 3) the whole spatial distribution of Changsha price is concentric oval like, and the core is traditional Wuyi business circle, and the direction of south- north has obvious tendency to proliferate; 4) the house price variance is regionally significant, and isolines in second ring are dense, while Hexi's(west district to Xiangjiang river) isolines are much sparser than that of Hedong(east district to Xiangjiang river), so its price is more slowly decaying.