However, if you are doing a DMRG simulation, you need to put in the parameter START_SWEEP and set it to the correct number. For example, if at first you ran with 32 states kept for 10 sweeps, you would then set the START_SWEEP parameter to 11 and SWEEPS to like 20. You could also change the number of states kept at this point. If you don't set the START_SWEEP parameter, the simulation will start all over. -----Original Message----- From: comp-phys-alps-users-bounces@phys.ethz.ch on behalf of Matthias Troyer Sent: Wed 1/28/2009 10:31 AM To: comp-phys-alps-users@phys.ethz.ch Subject: Re: [ALPS-users] Extending a simulation On 27 Jan 2009, at 03:21, Jakub Zakrzewski wrote:

Hello,

I am fresh, just trying ALPS, so question may be trivial. How to "extend a simulation" in QMC (worm). Info in wiki: http://alps.comp-phys.org/mediawiki-1.9.3/index.php/Running says to increase the amount of sweeps (which I did) and set a simulation status ti running from finished. But where to set it? Example -- ... is missing...

If your parameters are called parms, then parameter2xml creates a file parms.in.xml, and the simulation creates a file parms.out.xml . The latter is the file to edit. Matthias

Sorry to write so much, but I think I fixed it. I just needed to update some components in cygwin. -----Original Message----- From: comp-phys-alps-users-bounces@phys.ethz.ch on behalf of Justin David Peel Sent: Wed 1/28/2009 6:33 PM To: comp-phys-alps-users@phys.ethz.ch Subject: Errors during DMRG simulations I have been getting some errors during DMRG calculations lately. I think the only two that I have seen are: *** ERROR 1: Lanczos could not open vectors.dat and *** ERROR: Could not open file system.dat I'm using cygwin to run ALPS 1.3.3. Any ideas for what is causing these errors? The calculations continue in spite of these errors. Should I be worried about these errors causing inaccuracies in my calculations? Thanks, Justin

Dear Jakub, Although extending your simulation is possible by rewriting XML files by hand, I would recommend you to set up a new simulation with larger value of SWEEPS and run it from the first. Indeed, if you want to reduce the errorbar significantly (say, by factor 10), then you have to increase SWEEPS at least by factor 10^2=100. In such a situation, the contribution from the previous run to the final statistics (as well as the total CPU time) is negligible. Thus practically you will lose nothing by re-running the simulation from the first. Best, Synge From: Matthias Troyer <troyer@phys.ethz.ch> Date: Wed, 28 Jan 2009 09:31:00 -0800

On 27 Jan 2009, at 03:21, Jakub Zakrzewski wrote:

...

Hello,

I am fresh, just trying ALPS, so question may be trivial. How to "extend a simulation" in QMC (worm). Info in wiki: http://alps.comp-phys.org/mediawiki-1.9.3/index.php/Runningsays to increase the amount of sweeps (which I did) and set a simulation status ti running from finished. But where to set it? Example -- ... is missing...

If your parameters are called parms, then parameter2xml creates a file parms.in.xml, and the simulation creates a file parms.out.xml . The latter is the file to edit.

Matthias