Code used for my masters thesis

Physarum polycephalum a.k.a slime mould

Project Layout

• src/ contains the actual implementation of stuff
• requirements.txt packages required to make the code work
• matplotlibrc defaults for plotting diagrams with Matplotlib

The individual source files have documentation on there own.

Installation and Dependencies

The code should work with Python 3.6 or above. Some distributions do not include Tkinter (only required for draw.py) or venv in their python packages, so you may have to install them separately with your package manager.

I recommend using venv (included since 3.3) to manage the dependencies in an isolated environment.

$ cd <project folder>
$ python3 -m venv master-convert
$ source master-convert/bin/activate
$ pip3 install --upgrade pip
(master-convert) $ pip3 install -r requirements.txt

Windows

To run this project on Microsoft Windows I recommend installing WinPython. You will need to install the following packages (through the WinPython Control Panel). Depending your specific versions of Python and WinPython more additional packages might be required.

• numpy+mkl
• opencv-python

Example

As a basic example the following shows how to go from the MATLAB data to clustered graphs. It assumes the venv is set up and activated.

(master-convert) $ python src/convert_matlab.py --output-dir results/converted/matlab/ data/matlab/2011_09_14_crop_1/
[...]
(master-convert) $ python src/cluster_ahead.py --output-dir results/clustered/matlab_ahead_s2 --look-ahead-steps 2 results/converted/matlab/2011_09_14_crop_1/
[...]
(master-convert) $ python src/draw.py results/clustered/matlab_ahead_s2/2011_09_14_crop_1/
[...]

Drawing

draw.py can navigate through a folder containing multiple graphs using the arrow keys. If the crop you loaded provides cluster information, you can track them by clicking on a node and than navigating.

Pressing t while viewing a picture will output TikZ-Code into the figures-folder (this does not include background pictures).

Word Definitions

• session: experiment run on one Petri dish, can contain one or more crops
• crop: data gathered from a single drop placed on Petri dish
• picture: a single snapshot capturing the state of the slime mould
• graph: network constructed from a picture