active-matter

Multi agent reinforcement learning for intelligent active matter

Python data analysis package for active and soft matter simulations

Active Matter simulation

A simulation framework for nonequilibrium statistical physics

Simulations of flocking behaviour in active agents, using a simple Viscek model

Code developed for a research project at the University of British Columbia focusing on a simple model of active particles, supervised by Jörg Rottler.

Sedimention of active Brownian particles in a box. Codes for numerical simulations and numerical solution of PDEs.

BARCODE (Biomaterial Activity Readouts to Categorize, Optimize, Design and Engineer

Code developped for a PhD project at the Université de Montpellier, supervised by Ludovic Berthier (Montpellier) and Robert Jack (Cambridge).

Inverse reinforcement learning on real trajectories of active microscopic swimmers (starting with C. elegans chemotaxis), to infer the reward function they behave as if optimizing — MaxEnt IRL, Deep MaxEnt IRL, AIRL.

Code developed for a research project at the University of Cambridge, supervised by Robert Jack.

Code to simulate systems of self-propelled particles with friction (self-propulsion directions unchanging) with Newton's equations of motion, numerically solved with the velocity-Verlet-scheme.

This study uses SPH-validated simulations to explore microbead dynamics under acoustic levitation and convective heat transfer. It highlights unique motion patterns from acoustic-convective interactions, offering insights into particle-fluid behavior in acoustically driven systems.

3D starling-murmuration simulator: the Pearce et al. (2014) projection model (true spherical-cap occlusion) and topological Reynolds boids, runtime-switchable, with real-time collective-behaviour metrics (order parameter, opacity Θ/Θ′) and GPU-instanced ModernGL rendering. Grounded in three founding papers.

Classifies the dynamical state of a simulated 3D non-confluent tissue (liquid, glass, viscosity-saturated, crystal) from a single trajectory, using a random forest and 1-D CNN over 65 dynamical and structural descriptors.

Python scripts for active matter analysis

This document describes the end-to-end implementation pipeline for the ware- house optimization system, from raw order data ingestion to production deploy- ment and continuous monitoring. The pipeline is designed for: • Modularity• Scalability: Handles 10K+ orders/day across multiple warehouses

Simulation of Active Brownian Particles with Graph Neural Networks

Active Matter & Particle Dynamics Simulation (Vicsek Model) using JAX and 2D FFT Density Power Spectrum Analysis.

Phystem is a software designed to assist in the construction and exploration of physical systems. Additionally, the repository contains pre-implemented physical systems that can be used by users interested in exploration.