Geographically weighted (GW) method is a type of spatial statistical framework. GW methods have been developed to tackle spatial heterogeneity in data, with a kernel that moves across geographical space. The GW method applies to a wide range of statistical analysis methods to explore the local geographical characteristics of data and its relationships in bivariate and multivariate data analysis. GW methods currently include (generalized) linear regression, summary statistics, and principal components analysis. They have further potentials to be extended to any statistical methods. To discuss future directions of GW method developments, we reviewed previous works regarding the state-of-art GW methods and available software and tools. As its customization is flexible, the GW method is feasible for any spatial phenomenon in cases where spatial heterogeneity is to be considered.
