Eulerian Parameter Inference Code

We recommend to use a virtual environment with a python 3.11 distribution to reproduce our results. Using anaconda and the environment file eulerpi_env.yml, the working virtual environment is set up through the prompt

conda env create -f eulerpi_env.yml

The environment file eulerpi_env.yml contains a human-readable list of all required dependencies. Consequently, these dependencies can also be installed manually.

In addition to this README and the environment file, the downloaded .zip file contains 5 individual python files and a Data directory that includes all data used for inference. Three of the files are named *plot.py and can be used to reproduce the plots of our article for one respective example model:

Figure 2 subfigures can be reproduced using temperature_plot.py
Figure 3 subfigures can be reproduced using corona_plot.py
Figure 4 subfigures can be reproduced using heat_plot.py

To this end, the files contain several function definitions and an executable part at the end. All subfigures produced by executing the *plot.py files are saved in the Figures folder. The Applications folder contains the inference results as .csv files.

Please note, that only the scientific plots can be reproduced with this code. The digital drawings were designed differently.

Two of the files rely on models imported from the eulerpi python package. In addition, the two files ending in *model.py specify additional models we used to produce the images. We employ those for the heat diffusion equation model and to also produce simulation results for validation data in the COVID-19 model. Each of the files is individually commented.

On an ordinary laptop, temperature_plot.py can be executed in approximately 1 minute of runtime. The same machine takes less than 3 minutes to execute corona_plot.py and less than 10 minutes to execute heat_plot.py. The three named scripts produce all scientific subfigures of Figure 2,3, and 4 of the manuscript.

Version History