Reasoning Symbolic Regressor

LLM-enhanced symbolic regression: A reasoning-driven AI that refines equations using structured feedback and adaptive learning.

🚀 Overview

ReasoningSymbolicRegressor integrates symbolic regression with LLM-powered reasoning, allowing AI to not only search for equations but also self-correct and refine them through structured feedback.

📦 Installation

You can install this package via PyPI:

pip install reasoning-symbolic-regressor

🔧 Usage

Note: Make sure your OPENAI_API_KEY is set.

from reasoning_symbolic_regressor import ReasoningSymbolicRegressor

# Initialize the AI reasoning-driven symbolic regressor
regressor = ReasoningSymbolicRegressor(debug=True)

# Fit the model to data
regressor.fit(X, y)

✨ Features

✅ LLM-Guided Exploration: Dynamically adjusts search parameters using AI reasoning.
✅ Self-Repairing Feedback: Detects errors in PySR configurations and prompts LLM to correct them.
✅ Iterative Refinement: Improves equations over multiple guided cycles.
✅ Early Stopping: Terminates when the LLM determines the correct equation has been found.

🛠️ Development

To contribute or modify the project, clone the repository and install dependencies:

git clone https://github.com/sidu/ReasoningSymbolicRegressor.git
cd ReasoningSymbolicRegressor
pip install -r requirements.txt

🧪 Testing

To run the tests, use the following command:

pip install -e .
python tests/test_gravitation.py

📜 License

This project is licensed under the MIT License.

🌟 Acknowledgments

Inspired by symbolic regression, LLM reasoning, and adaptive AI systems.