This project explores simulation-based inference (SBI) for solving the 3-body problem — a classic problem in physics involving the motion of three celestial bodies under mutual gravitational influence.
Using simulation data and neural posterior estimation, the goal is to approximate the posterior distribution over parameters given observational data, enabling inference in cases where the likelihood is intractable.
- Simulate 3-body trajectories
- Visualize trajectories
- Use SBI to infer initial conditions
- Evaluate posterior estimation
git clone https://github.com/chandanreddy10/Simulation-based-inference.git
cd Simulation-based-inferencepython -m venv venv
source venv/bin/activate pip install -r requirements.txt