Starter kit

Starter kit#

To start using the model and its dependencies, multiple scripts and tool should be used.

  1. Get the model

    • Clone the git repository from GitHub.

    • Install your environment using envs folder (latest data were produced with envs/environment.aws_r6a.12xlarge.yaml). A detailed installation guide is available in the PyPSA-Eur documentation.

    • Optimisation has been done using Gurobi (commercial solver). HiGHS is a good open source alternative. Other alternatives are listed in the PyPSA-Eur documentation.

  2. Run the model

    • Use snakemake using the following command

    snakemake -call all --configfile config/config.veka.yaml -n

    • Model can be ran locally or on a remote server to improve performances (should work on a AWS EC2 r6a.12xlarge).

    • Edit the configurations files to improve the scenarios : config/config.veka.yaml and config/scenarios.veka.yaml.

  3. Extract data from the PyPSA-Eur outputs

    • Copy desired files from results and resources folders to analysis/{run} folder. If you use a remote server, use SFTP to download data from the server.

    • Configure the correct run versions in scripts/graph_extraction_main.py for scenarios, sensitivities and reference runs.

    • Run the data extraction pipeline (locally).

    • The output data (analysis/{run}/graph_extraction_st) are created in analysis folder and should be copied in the app/assets/data folder if you want to add them to the streamlit.

  4. Ship the data in streamlit

    • Streamlit platform is free to use when the data are available on Github publicly. Currently our website is linked to main-v10 branch on the Github of Climact.

    • To use the new data, reconfigure app/st_common.py with the new runs paths (change scenario_dict values).

    • Merge the updated data with the GitHub branch main-v10 to ship the data online.

    • Streamlit will automatically update the webpage. If you want to be sure that cached data are updated, you can reboot the app on the Streamlit dashboard.

    • The webpage can be debugged locally using:

    streamlit run app/VEKA_2050+.py

    • Since Streamlit can also be run as a container, we have added app/Dockerfile_demo as an example.

  5. The documentation is shared online through Readthedocs.

    • Pushing new updates on the main-v10 branch on GitHub will also automatically update the page.

    • Documentation is stored in the doc folder.

    • The documentation can be tested locally using:

    cd doc
make clean
make html
open _build/html/index.html

    • The documentation is also configured to generate a PDF file using:

    cd doc
make clean
# Need twice for LaTeX references
make latexpdf; make latexpdf
open _build/latex/PyPSA-Eur.pdf