An Integrated Framework of Residential Community Construction year and Building Height Analysis

This repository contains the code and sample data for "Spatiotemporal Patterns and Influencing Factors of Residential Building Height in Beijing". Files are numbered to reflect the analytical workflow: (1) Landsat time-series extraction via GEE, (2) Random Forest binary classification of pre/post-2000 communities, (3) classifier validation, (4) temporal segmentation for construction year identification, (5) temporal segmentation validation, and (6) XGBoost-SHAP influencing factor analysis. Place data files in ./data/ before running scripts. Required Python packages: pandas, numpy, matplotlib, scikit-learn, xgboost, shap, geopandas, statsmodels, scipy, seaborn, joblib.