Formal Models of Reflective Equilibrium - Technical Report

This repository contains the technical report "Assessing a Formal Model of Reflective Equilibrium" for the SNSF-DFG project "Formal Models of Reflective Equilibrium - How Far Does Reflective Equilibrium Take Us?".

You can read this report in the browser (https://re-models.github.io/re-technical-report/) or download it as pdf file (https://doi.org/10.5281/zenodo.13294165).

This repository has the following structure:

• data: Contains all data files on which this report is based.
• notebooks: Jupyter notebooks to reproduce the data analysis (see below).
• report: Quarto source files for this report.
• output: Quarto generated output.
• re-technical-report: A small Python package that provides some helper functions for data generation and analysis.

Reproducing the Results

All findings and the underlying data can be reproduced by using the Python implementation of the model. For each chapter you will find here a Jupyter notebook whose execution produces all analysis results. You can execute them on Colab or locally by cloning this repository.

Using Colab

Simply click on the Colab badge () in the notebooks. Note, however, that it is not possible to use the complete data set on Colab. Instead, you are confined to using the subdata set TINY (see below).

Running the Notebooks locally

You can execute the notebooks locally by proceeding along the following steps:

1. Install python together with JupyterLab (https://jupyter.org/install)
2. Git-clone the repository (with git clone git@github.com:re-models/re-technical-report.git).
3. Open notebooks in JupyterLab and execute the cells.

By default, the notebooks use a convenience function to import one of the three data sets by specifiying the data set's name (e.g., re_data_by_name(data_name="TINY")). The function handles data retrieval in dependence of whether the notebook is executed on Colab or locally.

Alternatively, you can load the data by specifying the directory and file name of the data file in the following way:

re_data = load_re_data(data_dir=path.join(Path(getcwd()).parent.absolute(), 'data'), 
                       data_file_name='re_data_tau_alpha_mini.tar.gz')

Data

The data that the model produced can be found here. We provide three different data sets:

1. The FULL data set (file name re_data_tau_alpha.tar.gz, extracted $\sim 12GB$): The full data set that is described and used in the report (👉 https://re-models.github.io/re-technical-report/intro.html#ensemble-description)
2. The MINI data set (file name re_data_tau_alpha_mini.tar.gz, extracted $\sim 800MB$): A subset of the full data set based on $20$ dialectical structures.
3. The TINY data set (file name re_data_tau_alpha_tiny.csv, $\sim 12MB$): A subset of the full data set based on $2$ dialectical structures.

Note that using the full data set might, depending on the available RAM, overload your computational resources.

Citing

BibTex citation:

@article{freivogel_assessing_2024,
  title = {Assessing a {{Formal Model}} of {{Reflective Equilibrium}}},
  author = {Freivogel, Andreas and Cacean, Sebastian},
  year = {2024},
  month = august,
  doi = {10.5281/zenodo.13294165},
  langid = {english},
  url = {https://re-models.github.io/re-technical-report/},
}

For attribution, please cite this works as, for instance:

Freivogel, A., & Cacean, S. (2024). Assessing a Formal Model of Reflective Equilibrium. https://doi.org/10.5281/zenodo.13294165

Credits

Earlier versions of this report were discussed on several occasions with all members of the research project 'How far does Reflective Equilibrium Take us? Investigating the Power of a Philosophical Method' (SNSF grant 182854 and German Research Foundation grant 412679086). We thank, in particular, Claus Beisbart, Gregor Betz, Georg Brun, Alexander Koch and Richard Lohse for their helpful comments, which helped to improve this report considerably. Finally, the authors acknowledge support by the state of Baden-Württemberg through the joint high-performance computer system bwHPC.

---

This work is licensed under a Creative Commons Attribution 4.0 International License.