gencrt 0.1.0

A package implementing the generalized Chinese remainder theorem (CRT) algorithm.

gencrt

gencrt is a Python library that implements the Generalized Chinese Remainder Theorem (CRT) algorithm. It enables solving systems of modular equations efficiently and robustly, making it a valuable tool for applications in number theory, cryptography, and computational mathematics.

Features

• Solve modular equations using the generalized CRT algorithm.
• Support for multiple congruences of the form x ≡ a_i (mod n_i).
• Includes utility functions for:
  • Extended Euclidean Algorithm to compute the greatest common divisor (GCD) and Bézout coefficients.
  • Validation of moduli and congruences for correctness.

Installation

To use gencrt in your project, you can install it via pip:

pip install gencrt

Usage

Importing the Library

from gencrt import crt, extended_gcd

Solving Modular Equations with generalized CRT

To solve a system of modular equations:

x ≡ a1 (mod n1)
x ≡ a2 (mod n2)
x ≡ a3 (mod n3)

from gencrt import crt

# Define the congruences as a list of (a_i, n_i)
congruences = [(2, 3), (3, 4), (1, 5)]

# Solve using the generalized CRT
solution = crt(congruences)

print(f"The solution is x ≡ {solution} (mod product of moduli)")

Extended Euclidean Algorithm

Compute the GCD and Bézout coefficients:

from gencrt import extended_gcd

a, b = 56, 15
gcd, x, y = extended_gcd(a, b)

print(f"GCD: {gcd}, x: {x}, y: {y} (Bézout coefficients)")

API Reference

crt(congruences: Iterable)

• Description: Solves a system of modular equations using the generalized CRT.
• Parameters:
  • congruences: An iterable of tuples (a_i, n_i).
• Returns: Integer solution x modulo the product of moduli, or None if no solution exists.

extended_gcd(a: int, b: int)

• Description: Computes the GCD of two integers and Bézout coefficients.
• Parameters:
  • a, b: Integers.
• Returns: A tuple (GCD, x, y).

License

This library is licensed under the MIT License. See LICENSE for more details.

---

Author: Simon Ljungbeck Date: December 2024

For issues or contributions, visit the GitHub repository.