Analysing Landscape Evolution Models with TDA and vineyards

This is the code for "Using topological data analysis and persistence vineyards as a descriptor for evolving landscapes".

Dependencies:

To install the required conda packages to use the environment.yml file included.

conda env create --file environment.yml
conda activate LEMvineyard

There are some dependencies which are not available on conda. A custom fork of Ripser.py is required, maintained by Dr. Gillian Grindstaff. To install, ensure first the conda environment from the previous command is active. Then run the following:

git clone https://github.com/ggrindstaff/ripser.py.git
cd ripser.py
pip install -e .

To run the Dionysus vineyards code, Dionysus 1 (Not version two!), is required. Specifically, once Dionysus is built, the file /path/to/dionysus/build/examples/pl-functions/pl-vineyard must be added to PATH.

Running Guide:

First LEM must be generated. This process is random but results will be similar each run given initial conditions. To do this run generate_LEM.py.

The code to generate the vineyards and draw the plots can then be run inside landscape_statistics.ipynb