Convergent divergent unit

Ilias Rentzeperis, Steeve Laquitaine

Please cite the following paper:

@article{rentzeperis2022adaptive,
  title={Adaptive rewiring of random neural networks generates convergent--divergent​ units},
  author={Rentzeperis, Ilias and Laquitaine, Steeve and van Leeuwen, Cees},
  journal={Communications in Nonlinear Science and Numerical Simulation},
  volume={107},
  pages={106135},
  year={2022},
  publisher={Elsevier}
}

Prerequisites

• Conda must be installed

Setup

Move to your project’s root directory.

conda create -n dgr python==3.7      # create dgr virtual environment  
conda activate dgr
pip install -r src/requirements.txt  # install requirements.txt 
ipython kernel install --name dgr    # create jupyter kernel for dgr

Run

1. Create raw directed graphs data (can take up to 2 days)

Open and run all cells of 1RunStoreDigraphs.ipynb. After re-creating the raw data you can either:

1. execute the analyses and plot the figures (1)

2. or directly plot the figures from the pre-stored analyses results.

3. Run analyses on raw data and create figures:

python -m main --run figure2
python -m main --run figure3
python -m main --run figure4
python -m main --run figure5
python -m main --run figure6
python -m main --run figureS3
python -m main --run figureS4

3. Or just load stored intermediate analyses and quickly re-create figures:

python -m main --load figure2
python -m main --load figure3
python -m main --load figure4
python -m main --load figure5
python -m main --load figure6
python -m main --load figureS3
python -m main --load figureS4