Network model of simple societies inspired by Joseph Tainter's theory of the collapse of complex societies.
Tainter, J. A. (1988).The collapse of complex societies.
New studies in archaeology. Cambridge: CambridgeUniversity Press
Obtain the code and change into directory
git clone [email protected]:flo-schu/tainter
cd tainter
Create environment, activate it and install model package. the [datalad] option installs the datalad packages which are necessary to download the dataset.
conda create -n tainter
conda activate tainter
conda install python=3.9
pip install -e .[datalad]
Install the pre-simulated dataset from the open-science foundation (osf.io) into the folder_publication data
datalad clone https://osf.io/u897c/ publication_data
You should be done. All anylses from the paper can be conducted by executing
the jupyter notebook scripts/analysis.ipynb
jupyter lab scripts/analysis.ipynb
The model is an agent-based network model written in Python. The main files are included under scripts/model.
to execute the model code, navigate (cd) into scripts directory. From here all scripts in the main directory can be executed.
Prefixes correspond to figures displayed in the publication
The files macroscopic_approximation.py and stochastic_model.py are very basic scripts which execute the main modules of the code.
data for f4 was generated on a high perfomance cluster. The necessary files for this computation are found under scripts/cluster. These files need to be slightly adapted to suit file structure. Simulations can be carried out on a normal computer if the resolution of parameters is reduced.
Several analyses can be performed with the model code.
-
stochastic simulations of the network with various paramter settings. For details on parameters see the inline documentation in model/main.py
-
analytic calculations with equations derived in publication
-
several plots to illustrate the results in publication