Reproducibility Package for Feature Reduction and the Stability of Diabetes Prediction on the Pima Dataset

This reproducibility package accompanies the study “Feature Reduction and the Stability of Diabetes Prediction on the Pima Dataset: A Repeated, Leakage-Safe Cross-Validation Study.” It contains the reconstructed analysis code, archived and reproduced results, revised statistical comparisons, sensitivity analyses, and figure files. The README documents the software environment, the SHA-256 checksum of the analysis CSV, and the steps required to reproduce the results.