Pure-Python symbolic regression that rediscovered Kepler's law from 8 data point

- Symbolic regression is a machine-learning technique that searches for a clean mathematical formula to fit data (e.g., y = a·x² + b), unlike neural networks which produce black-box models. - Kepler's third law (P² ∝ a³): the square of a planet's orbital period equals the cube of its distance from the Sun. Discovered in the early 1600s from observational data, it paved the way for Newton's gravity. - GP-Elite is a pure-Python symbolic regression library using genetic programming — evolving formulas by randomly combining operators (+, −, ×, ÷, powers), selecting the best fits, and mutating/recombining them over generations. - Its demo: given just 8 data points (planetary distances and orbital periods), it rediscovered the exact Kepler law formula without any hints about the relationship's form. - "Pure-Python" means no heavy dependencies (no NumPy, no PyTorch), making it easy to run and study.