Open and Reproducible Data Analysis for the Modern Eddington Experiment

The Modern Eddington Experiment (MEE) aims to improve the precision and accessibility of this classic test of general relativity. While weather compromised the precision of measurements from the April 8, 2024 eclipse, the MEE2024 report identified its open-source analysis software as the effort's greatest success. Yet the pipeline required manual GUI interaction at each stage, and the data (450 GB) resided only on personal drives. We developed infrastructure that makes MEE analysis genuinely open and reproducible: TOML configuration files that capture every parameter under version control; a Python API that executes the complete pipeline — image stacking, plate-solving, distortion fitting, refraction correction, and deflection measurement — without manual interaction; and MEE2024 data hosted on Kaggle, enabling cloud execution accessible to collaborators worldwide. To demonstrate these capabilities, we present a systematic study of how zenith field and exposure affect the precision of the gravitational deflection measurement. These data-driven results provide quantitative guidance for observing protocols in upcoming expeditions to Spain (2026), Africa (2027), and Australia (2028).