pinn-kit 0.1.1

A toolkit for Physics-Informed Neural Networks

PINN-kit

A toolkit for Physics-Informed Neural Networks (PINNs). This package provides tools and utilities for implementing and training physics-informed neural networks for solving differential equations.

Features

• Easy-to-use interface for defining physics-informed neural networks
• Support for various types of differential equations
• Domain handling utilities
• Training and evaluation tools

Installation

PINN-kit supports macOS, Linux, and Windows. You can install it using pip, which will also automatically install all required dependencies, including torch.

Create a Virtual Environment (Recommended)

It is recommended to use a virtual environment with Python 3.12 to avoid dependency conflicts.

macOS & Linux

python3.12 -m venv venv
source venv/bin/activate

Windows

Open Command Prompt or PowerShell and run:

python3.12 -m venv venv
venv\Scripts\activate

Install PINN-kit

macOS & Linux

pip install pinn-kit

Windows

Open Command Prompt or PowerShell and run:

pip install pinn-kit

Note: There is no need to install torch separately; it will be installed automatically with pinn-kit.

PINN-kit is easy to use

from pinn_kit import PINN, Domain

# Create a domain
domain = Domain(...)

# Initialize a PINN
pinn = PINN(...)

# Define the loss terms
def loss():
    return loss_function

# Train the network
pinn.train_model(...)

Documentation

For detailed documentation, please visit our documentation page.

License

This project is licensed under the MIT License - see the LICENSE file for details.