Dataset for: Machine Learning vs Langmuir – A Multioutput XGBoost Regressor Better Captures Soil Phosphorus Adsorption Dynamics

This dataset supports the study “Machine Learning vs Langmuir – A Multioutput XGBoost Regressor Better Captures Soil Phosphorus Adsorption Dynamics.” It contains soil physicochemical data and corresponding phosphorus adsorption measurements used to compare traditional Langmuir isotherm fitting with machine learning approaches for modeling phosphorus retention in agricultural soils. The machine learning model, specifically a multi-output XGBoost regressor, was trained to predict phosphorus adsorption across multiple equilibrium concentrations simultaneously, using features such as soil texture, pH, organic matter, extractable nutrients, and cation exchange capacity. The repository includes: An xlsx file with 147 soil samples and corresponding measured and predicted adsorption values at Ce = 1, 2, 4, 6, and 10 mg/L An xlsx file with 10493 soil samples used for comparing the Langmuir isotherms against the multiouput XGBoost model These datasets are intended to facilitate reproducibility, future benchmarking, and meta-analysis on data-driven phosphorus modeling in soils.