Hi,
I'd like to use the classical Monte Carlo code to study the properties of the 'Heisenberg-XY' model, where the symmetry is a product of O(3) and O(2), i.e. the degrees of freedom at each site can be represented by (n, \theta) where n is a vector and \theta is an independent angle. Is it feasible to do this by editing the current mc/spin files alone? From inspection it seems that, at the very least, I would need to change parts of spinmc_factory.C and create something similar to the file on.h, but allowing for O(a) * O(b) symmetry.
Thanks for any advice you can give,
Andrew
Hi, Andrew,
On Sat, Jan 16, 2010 at 5:45 AM, Andrew James andyjaj@gmail.com wrote:
Is it feasible to do this by editing the current mc/spin files alone? From inspection it seems that, at the very least, I would need to change parts of spinmc_factory.C and create something similar to the file on.h, but allowing for O(a) * O(b) symmetry.
That's correct.
The simplified classical MC code for the Ising model can be found inside example/scheduler/ directory.
If you want only Metropolis algorithm, editing example application code may be easier than spinmc code.
Best wishes,
comp-phys-alps-users@lists.phys.ethz.ch