The above code uses the Metropolis–Hastings MCMC algorithm to simulate the 2D Ising Model. We first perform a series of measurements of various observables, namely -
Once we obtain the autocorrelation time
On obtaining the uncorrelated measurements for the observales, we calculate the errors in the observables and the derived physical quantities (
As of now, we have used a very naive version of the data-collapse method to evaluate the critical exponents without taking into account the Jackknife errors evaluated for our physical quantities. Anders Sandvik's document on Computational Studies of Quantum Spin Systems is an excellent resource on all things related to spin system simulations, including finite size scaling analysis and error analysis as well.