soil-spectroscopy

R package: Misc. Functions for Processing and Sample Selection of Spectroscopic Data

resemble is an R package for similarity-based modelling and local learning in spectroscopy. It provides tools for dissimilarity computation, nearest-neighbour search, memory-based learning, and spectral library optimisation (methods designed for large, heterogeneous spectral datasets where global models underperform)

R scripts for predicting soil organic carbon using soil spectral library from visible, near-infrared and shortwave-infrared (VNIR) and middle-infrared (MIR) using LASSO and PLS regression methods and the target-oriented cross-validation strategy.

A Python package for handling soil spectroscopy data, with a focus on the Open Soil Spectral Library (OSSL).

Prediction of Exchangeable Potassium in Soil through Mid-Infrared Spectroscopy and Deep Learning: from Prediction to Explainability, Albinet et al., 2022

Soil VIS-NIR reflectance spectra simulation based on generative model. PROSAIL model is integrated..

Functions to analyse mid infrared spectra of peat samples

Comparing different data preprocessing methods to predict soil organic carbon content on soil spectra features

Provides Scikit-Learn compatible transforms for spectroscopic data preprocessing.

R implementation of a Vis-NIR soil spectroscopy workflow for predicting soil properties using Principal Component Regression (PCR), Partial Least Squares Regression (PLSR), Cubist, Random Forest (RF), Support Vector Regression (SVM), Memory-Based Learning (MBL), Artificial Neural Networks (ANN) and Extreme Gradient Boosting (XGBoost) algorithms

Predicting sand, silt and clay from VisNIR soil spectra. Shows that the choice of validation costs 5× more accuracy than the choice of sensor.

Code and precomputed results for "Unconditional Flow Matching With Classifier-Based Pruning for Distribution-Aligned Soil Spectral Synthesis" (IEEE GRSL 2026). Generates synthetic LUCAS 2015 topsoil data with flow matching, pruned by a classifier to align distributions.

Generative deep learning (cWGAN-GP) for vis-NIR soil spectral library augmentation. MSc thesis code.