zap documentation

Zap is a Python library for for differentiable network optimization. Currently, it provides a modular, component-based approach for modeling electricity systems and congestion-controlled flow networks. Models can be solved using a GPU-accelerated ADMM solver, integrated into PyTorch-based deep learning models, or iteratively planned using gradient algorithms.

Note

This project is under active development.

• Electricity System Modeling

• Network Utility Maximization