A new test for random number generators: Schwinger-Dyson equations for the Ising model

TL;DR

This paper introduces a novel, efficient test based on Schwinger-Dyson equations for evaluating the quality of random number generators using the Ising model in two and three dimensions.

Contribution

It presents a new, cost-effective method leveraging Schwinger-Dyson equations to detect bias in random number generators within Ising model simulations.

Findings

The test effectively detects bias in random number generators.

It is computationally inexpensive when integrated into simulations.

The method is sensitive to biases in both 2D and 3D Ising models.

Abstract

We use a set of Schwinger-Dyson equations for the Ising Model to check several random number generators. For the model in two and three dimensions, it is shown that the equations are sensitive tests of bias originated by the random numbers. The method is almost costless in computer time when added to any simulation.