Dear all, while performing some simulations in BH model (DWA code), I encountered problems with convergence in the vicinity of the critical point for small temperature T and a large linear dimension L. What is the root cause for the complexity of the simulation in such a case? I seem to get a large systematic error (which looks non-random as a a function of hamiltonian coupling constants). Is there any way to improve convergence?
While searching output hdf5 files for additional clues, and the DWA documentation, I have encountered the error_convergence in output files. Could I ask what exactly it is? Is it related to the "binning analysis" which has been discussed throughout the alps website?
Is there is a build-in method for convergence analysis? that would be very helpful.
Best regards, Oskar A. Prośniak
Hi,
Loot at tutorial http://alps.comp-phys.org/mediawiki/index.php/ALPS_2_Tutorials:MC-01_Autocor... to understand critical slowing down. That is what seems to be happening. The required runtimes diverge with problem size close to a phase transition.
You need to be careful in that regime, start at high T and small L, and slowly increase L or lower T once you are sure that your data has converged.
Matthias
On Jan 22, 2019, at 11:57, Oskar Amadeusz Prośniak oskar.krak@gmail.com wrote:
Dear all, while performing some simulations in BH model (DWA code), I encountered problems with convergence in the vicinity of the critical point for small temperature T and a large linear dimension L. What is the root cause for the complexity of the simulation in such a case? I seem to get a large systematic error (which looks non-random as a a function of hamiltonian coupling constants). Is there any way to improve convergence?
While searching output hdf5 files for additional clues, and the DWA documentation, I have encountered the error_convergence in output files. Could I ask what exactly it is? Is it related to the "binning analysis" which has been discussed throughout the alps website?
Is there is a build-in method for convergence analysis? that would be very helpful.
Best regards, Oskar A. Prośniak
Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/
List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users
Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch.
comp-phys-alps-users@lists.phys.ethz.ch