This is the public repository maintained by the author team for the publication of the article "Demographic transition hinders climate change mitigation in the aging and shrinking Japanese residential sector".
Restricted by the data policy of the Japan government, the original data or survey cannot be shared publicly on the github, therefore only the preprocessed data are available in this repository. Please contact us for further information if you are interested in the original data.
You should get a properly installed Python3 environment ready before running the repository. It is highly recommended to start with miniconda3 or anaconda3 for the environment setup. Once the environment is ready, please run the following code to install the dependencies:
pip install -r requirements.txt
All the dependencies required in the modelling of our paper will be installed automatically. Then, ensuring the original data (if requested) is available under data-local/ before running the tests, you can follow up the steps:
- Step 1: get preprocessing done
python preprocess.py
It outputs, 1) data/feature-desc-<timestamp>.xlsx, 2) cache-preprocess-data-<timstamp>.xlsx
- Step 2: run clustering model
python cluster.py
It outputs, 1) data/variable-importance-figure-<timestamp>.png, 2) data/cache-clustered-<timestamp>.csv, 3) data/cache-scaled-data-<timestamp>.csv
- Step 3: output statistical results
python stats.py
It outputs, 1) data/img/dist-emission-by-<type>-<timestamp>.png images for the distribution of per capita emissions by types, 2) data/household-by-prefecture-cluster-<timestamp>.xlsx, 3) data/emission-by-fuel-<timestamp>.xlsx, 4) data/emission-by-usage-<timestamp>.xlsx, 5) data/figure2-emission-by-potential-<timestamp>.xlsx
- Step 4: projection data
python predict.py
It outputs, 1) data/est-pv-coef-<timestamp>.xlsx, 2) data/est-pv-coef-<timestamp>.xlsx, 3) data/figure3-projection-emission-<timestamp>.xlsx, 4) data/japan-pop-projection-cluster-<timestamp>.xlsx
Only the main data relevant to clustering and projection have been included in this repository. For the other relevant data and scripts, please contact us for details.
- Yi Wu, y_wu.21@ucl.ac.uk
- Yin Long, longyinutokyo@gmail.com>