Generalized spatial error for validation

This repository contains the R scripts to reproduce the analysis presented in the paper: "Generalized spatial error for validation" .These scripts implement the novel 'Generalized Spatial Error' (GSE) metric, a spatially-aware validation tool designed to overcome the limitations of traditional, non-spatial error metrics like RMSE in geospatial modeling.The code allows users to replicate the key findings of the paper, including:Pre-analysis of explanatory variables (VIF and correlation matrix; Figure 4).Training and baseline performance evaluation of nine machine learning models (Table 3).The complete end-to-end workflow for GSE calculation and the determination of the optimal analysis scale (Figure 8).Detailed instructions on software requirements, data preparation, and script execution order are provided in the README.md file. Users of this code are encouraged to cite the original publication.