Fig. 7: Results of cost factor screening.
From: Machine-learning-based reconstruction of Ming-dynasty defensive corridors in Yuxian
a Heatmap of the Spearman’s rank correlation coefficient matrix among cost factors, with colors ranging from light to dark corresponding to correlation coefficients varying from −1 to 1. The diagonal represents the perfect correlation of each variable with itself (ρ = 1), and the values denote the pairwise Spearman’s rank correlation coefficients of the factors; b Visualization diagram of correlation strength and significance. The colors of the rings also indicate the positive/negative nature and magnitude of Spearman’s ρ, the arc length of the sectors reflects the correlation strength, and the sizes of the fan-shaped areas inside the rings represent the significance of the correlation at the levels of p < 0.05, p < 0.01, and p < 0.001, respectively. Factor pairs with high |ρ| values are considered to have potential redundancy, which were prioritized for examination and elimination in the subsequent VIF analysis and cost factor screening.
