Introduction

The code in this repository was written to run the analysis done in [2] and create the pictures therein. It is the analysis of low-complexity socio-economic model (AYS-model) with respect to the TSM framework [1]. If you find any of the work interesting, please don't hesitate to contact me (Tim.Kittel@pik-potsdam.de).

[1] http://www.earth-syst-dynam.net/7/21/2016/esd-7-21-2016.html

[2] https://arxiv.org/abs/1706.04542

Setup

The code was tested with Python 3.5 under Ubuntu Xenial only. If you want to get it running on a different system, please contact me (Tim.Kittel@pik-potsdam.de).

The Code relies on the PyViability-library, please install it.

The easiest way to install everything is to copy and run the following lines

git clone https://github.com/timkittel/ays-model.git
git clone https://github.com/timkittel/PyViability.git
pip install matplotlib==1.5.3
cd PyViability
pip install -e .
cd ..

Note that it's important to have the correct version of matplotlib. If the version is too new, there are problems with the opacity of patches and the plots are not correctly shown.

auto-completion

The scripts all provide auto-completion of command line arguments via argcomplete (install with pip install argcomplete). If you want to use it, please run once

activate-global-python-argcomplete --dest=- >> ~/.bash_completion

Afterwards you should be able to use [Tab] for auto-completion.

Running the Code

Disclaimer I: To understand what the Code does, please (at least) skim through [2] first.

Disclaimer II: This code is prepared for scientific work, not a production where security and tight right management is important. The eval function is used for simplicity during command line evaluation ...

Disclaimer III: If you read the code, often w/W is used instead of the variable y/Y in [2]. We renamed it later, because Y is a common choice for economic production. The same goes for degrowth/dg instead of lowgrowth/lg.

In order to follow through the computations done for [2] and create the respective pictures, there are four steps to take:

1. create the "hair plots" Fig. 3 and Fig. 4 (a)-(c)
2. running the actual analysis of the model,
3. displaying the result as in Fig. 5, and
4. the bifurcation analysis, Fig. 6.

All of the scripts have been written with rather high verbosity in mind so the user is aware what's happening. Still, if you don't find it verbose enough, check out the -v flag.

Creation of the "hair plots"

To create the plots that show the dynamics of the model, Fig. 3 and 4 in [2], use the following lines

./ays_show.py default --no-boundary
./ays_show.py default
./ays_show.py degrowth
./ays_show.py energy-transformation

Running the TSM analysis on the model

Warning: Running the analysis takes a while (on my Laptop ~ 90 Minutes).

To run the actual TSM analysis, ays_tsm.py is used. It has an extensive help, please check ./ays_tsm.py -h. The analysis in the paper was finally done with:

./ays_tsm.py --dg --et --num 200 --boundaries both dg-et-both.out

The arguments stand for: --dg : use the low growth management option (see [2]), --et : use the energy transformation management option (see [2]), --num 200 : the grid should have 200 points per dimension, i.e. 200^3 in total, --boundaries both : use both boundaries (see [2]), and dg-et-both.out : file name of the output file.

Create the result plots

Warning: The plotting take quite a bit of memory, ~ 16 GB.

To plot the output, use ays_tsm_show.py. The surface of the region(s) is estimated using alpha shapes. As this process takes a while, the script will save a cache file to speed up the process the next time it's run. To create the two plots of Fig. 5 in [2], run

./ays_tsm_show.py dg-et-both.out  --paper -r s b -s shelter-backwater.jpg

and

./ays_tsm_show.py dg-et-both.out  --paper -r l -s current-state-in-lake.jpg

Please check the help message for details on the arguments.

Bifurcation analysis

The bifurcation analysis is basically not doable without access to an High-Performance Computer. I simply changed the corresponding parameter of the bifurcation analysis using the --set-parameter option of ays_tsm.py and ran a full analysis for all cases. Then use ays_tsm_bifurc_show.py to plot the results.

More can be done

There are more features for more extensive TSM analysis. Please check the help of the scripts or, if your interest won't be satisified with that, I am very happy (to try) to answer any question (Tim.Kittel@pik-potsdam.de).