Closed guozihuaa closed 6 years ago
Dear Zihu
You are right in saying that: X = ( < M^2 > - < M >^2 )/T
Note <M^2> is defined per monte carlo move. So <M^2> = M2/(mcSteps).
<M> = M1/mcsteps, so <M>^2 = (M1/(mcSteps)) * (M1/(mcSteps))
In the end, you make it per site by dividing by N*N.
So n1 = 1.0/(mcSteps * N * N);
and n2 = 1.0/( mcSteps * mcSteps * N * N),
Hope that helps for your comment on https://rajeshrinet.github.io/blog/2014/ising-model/.
I think Susceptibility should be calculated as:
X[tt] = (n1*M2 - n1^2*M1^2)*iT
because
X = (<M^2> - <M>^2)/T