Stanford-Mineral-X/DGSA: v1.0.0 — Initial Release

v1.0.0 — Initial Release

First public release of DGSA (Distance-based Generalized Sensitivity Analysis), a Python package for sensitivity analysis of geoscientific computer experiments.

Features

• K-medoids clustering on user-supplied distance matrices with reproducible random seeding
• Single parameter sensitivity via L1-norm and ASL (Achieved Significance Level) methods
• Conditional (two-way) parameter sensitivity to quantify asymmetric parameter interactions
• Visualization tools: Pareto plots, heatmaps, CDF plots, and MDS cluster plots using perceptually uniform colormaps (cmcrameri)

Example Dataset

Includes the Park et al. (2016) reservoir modeling dataset for reproducing published results.

Requirements

• Python >= 3.10
• Dependencies: numpy, pandas, matplotlib, scipy, scikit-learn, cmcrameri

References

• Fenwick et al. (2014), Mathematical Geosciences
• Park et al. (2016), Computers & Geosciences