wuyi0614 / UrbanClimate-ChinaHCE

The repo for the data and codes from the paper about China's HCE to be published on Urban Climate

UrbanClimate-ChinaHCE

The repo for the data and codes from the paper about China's HCE to be published on Urban Climate.

Outline

Edited on 27 Oct 2023

Notably, as accounting the carbon emission based on the survey is rather difficult, we use energy consumption instead for analysis.

Title: City-level household energy consumption typology and implications: a machine learning-based approach(pending)

• Abstract (dependent on findings) - YW
• Introduction & Literature Review
  • Cities account for a majority of energy consumption & household energy/lifestyle energy consumption
  • Understanding of Chinese cities' energy consumption and emissions
  • What approaches have been applied to explore HCEs and what findings regarding urban/rural HCEs
  • What gap exists in the current researches? -- highly dependent on a fixed range of factors, and lack of deep understanding of lifestyle and behaviors of Households
• Methodology and data description (& a graphic methodology)
  • Survey samples - YW & ZYX
  • Energy consumption and emission data processing and inequality analysis - YW
  • Machine learning approaches - ZYX
• Results
  • City-level energy consumption and emissions - YW
    • city-level energy consumption
    • city-level appliance ownership
  • HEC inequality by types and by regions - YW & ZYX
    • General inequality by types (essential/additional or electrified/fossil)
    • Regional inequality and their components
  • A machine learning-based HCE typology - ZYX
    • Typology analysis - LASSO + factor identification
    • Urban/rural typology analysis - clustering (attributes analysis)
• Discussion
  • Conclusion
  • Policy implication
  • Limitation
• Appendix
  • S1: Household energy consumption estimation
  • S2: Machine learning approach

Quick start and test

To start building the dataset, please follow the steps by,

• run build.py with raw data, i.e. CGSS-unprocessed-202302.xlsx, and specify the output file, e.g. vardata-<mmdd>.xlsx
• run check.py to add up vehicle and fuel data, and usually the output file has the same datafile name
• run merge.py to mapping household features with energy consumption data by ids
• run cluster.py to process data and do clustering experiments
• run typology.py to export the UI-friendly clustering summary
• run result_one.py to produce the first batch of figures
• run result_two.py to produce the figures of Lorenz and others