Code for: Data Fusion-Based Outlier Detection and Carbon Emission Estimation for China's Construction Industry

This repository contains the Python scripts accompanying the manuscript "Improving carbon emission estimates through data fusion: Evidence from China's construction industry". The code implements: (1) a multi-indicator data fusion framework for detecting and correcting outliers in provincial building material consumption data across 30 Chinese provinces (2004–2021), using cross-indicator validation, ratio stability analysis, and Huber robust regression; and (2) carbon emission calculation for the construction industry, comparing results from original and corrected data. See README.md for usage instructions and directory structure.