Design of a modular solver - Excel spreadsheet decision-support tool for calibrated nonlinear farm optimization in...

This study presents the design of a modular General Algebraic Modelling System (GAMS) solver integrated with an Excel spreadsheet as part of a decision-support system for implementing calibrated nonlinear farm optimization models under data-limited smallholder farm conditions. The study does not claim methodological innovation in nonlinear Positive Mathematical Programming or causal inference regarding farmer behavior. Instead, the optimization model is treated as a computational component embedded within a broader digital decision-support framework. The model system separates data input, scenario execution, model calibration, and output interpretation into modular components, enabling repeated simulations without requiring model modification. A baseline linear programming (LP) model is sequentially calibrated using dual values to recover implicit costs. The quadratic cost parameters are specified such that base-year replication matches the observed values exactly. The model includes automated looping routines that allow systematic price perturbation and supply response analysis. Validation employs base-year replication metrics, convexity checks, and solver optimality checks. Overall, the model provides a transferable framework that embeds nonlinear farm optimization into extension-relevant decision-support environments while maintaining computational transparency and reproducibility.•The design decouples the input data from model execution, enabling rapid simulations without interfering with the underlying GAMS code.•The calibration process is sequential, utilizing dual values to recover implicit costs, ensuring the model exactly replicates observed smallholder data behavior in data-limited smallholder farm environments.•The automated looping routines enable high-speed, systematic price sensitivity and supply response analysis, providing a transparent and reproducible architecture for agricultural extension optimisation advice and ex-ante measures.