Reproducibility package for station-level short-term passenger demand forecasting with lightweight machine-learning models

This record contains the anonymous reproducibility package for a station-level short-term public transport passenger-demand forecasting study. The package includes synthetic benchmark scripts and outputs, public MTA Subway Hourly Ridership external-validation scripts and outputs, processed analysis-ready data, prediction-level results, statistical-test outputs, figures, documentation, and environment files. The package is provided to support independent reproduction of the reported synthetic and public-data validation experiments. The repository includes a license file. Code components are distributed under the MIT License, while generated data, documentation, figures, and result tables are made available under CC BY 4.0 unless otherwise stated. Third-party public data sources are documented in the included data-source notes.