Welcome to InvODE’s Documentation!

InvODE is a Python library for flexible and interpretable inference of ODE-based models from time series data. Whether you’re modeling epidemics, enzyme kinetics, population dynamics, or financial systems — InvODE is built to help you fit, refine, and understand your dynamical models.

✨ Key Features

• 🔍 Flexible optimization for parameters, initial conditions, and latent states.

• 🧩 Custom loss functions, and regularization terms.

• ⚙️ Solver-agnostic design with compatibility across odeint, solve_ivp or other ecosystems.

• 🧪 Support for replicates, noisy data, and identifiability checks.

• 🌐 Domain-agnostic: Works for models in biology, chemistry, physics, and more.

• 📦 Ready for research and deployment — ideal for rapid experimentation and rigorous validation.

😺 Visit our GitHub

For the latest updates, source code, and to contribute, visit our GitHub repository: RaunakDey/invode

🚀 Get Started

Quickstart

• Quickstart

📚 Learn & Explore

• Simple Tutorials
  • Fitting vs Non-Fitting initial condition
  • Fitting replicates
  • Adding constraints
• Case Studies
  • Lotka–Volterra Case Study
  • Chemical Kinetics Case Study
  • Phage One-Step Growth Case Study

🛠 API Reference

• API Reference

🤝 Contributing & Community

We welcome contributions! Whether it’s fixing bugs, writing documentation, or adding features, we’d love your help.

• Contributing Guide

—

InvODE is developed with ❤️ by scientists who care for transparency and open source ecosystem. If you use this package in your research, please consider citing us (citation coming soon!).