This repository holds code related to "Synchronization in phase-coupled Kuramoto oscillator networks with axonal delay and synaptic plasticity" https://journals.aps.org/pre/abstract/10.1103/PhysRevE.89.032906. It simulates emergent behavior of coupled oscillator networks with learning rules and time delays that more closely mimic the interactions between neurons in the brain than in traditional machine learning "neuronal network" based approaches.
It is currently split into 3 directories, each containing code for one language:
This folder contains a copy of the script (covering both the model and some plotting), originally written circa 2013 for my undergraduate honors thesis and the publication above. It is a "procedure file" for a program called Igor Pro which enabled easy and powerful visualization at the time but is, unfortunately, both locked-down/proprietary and obscure (relative to alternatives like MATLAB and Python). Those aspects limited the potential use of the code and direct exploration of the models by others. I also didn't know how to use git
, python
, etc. at the time, so the file here is just copied from surviving copies I had in an old backup. It is here just for the historical record.
In August 2022, I translated the Igor Pro modeling code to the Rust programming language. Doing so provides:
- greater accessibility of the model since Rust is entirely open source (and I can provide pre-compiled binaries),
- much faster execution with modern optimizations and C-like performance,
- a project for me to learn Rust.
To build the binary, one should install the Rust toolchain and run
cargo build --release
inside the kuramoto-rust
directory. That will create an optimized binary under the target/release
subdirectory. The CLI parameters available for the program can then be listed via,
./release/kuramoto_rust -h
and the kuramoto_rust
binary can then be copied anywhere on your system. Documentation isn't fully complete currently but is available under target/doc/kuramoto_rust
when built with cargo doc
.
This folder contains some simple python scripts for plotting and animating the resulting data files from the rust program. It has preliminary examples for the 1D and 2D cases.