[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ESPResSo-users] stopping the simulation
From: |
Peter Kosovan |
Subject: |
Re: [ESPResSo-users] stopping the simulation |
Date: |
Wed, 11 Jul 2012 17:07:30 +0200 (CEST) |
User-agent: |
Alpine 2.00 (LNX 1167 2008-08-23) |
Hello Tarun,
You can get full answer to your question only if you provide details of your
acutal simulation setup, e.g. by sending your tcl script.
I assume you are using the Langevin integrator which includes random forces. If
you do not initialize your random number generator, the default seed is used
when you start (or re-start) your simulation. To obtain exactly the same
trajectory from a signle run and from a simulation where you save/restore the
configuration, you need to save also the status of random number generators on
each node and when restoring, you also need to restore the rng status.
On the other hand, if you are not studying time-dependent quantities, it should
not matter. The trajectories between the single run and the saved/restored will
be different but will have the same statistical properties. If you re-init your
random number generators 6 times for the whole simulation, its influence should
be well below the statistical error of the simulation.
Regards,
peter
--
Peter Košovan
Institute of Computational Physics
University of Stuttgart
Pfaffenwaldring 27
70569 Stuttgart
Germany
www.icp.uni-stuttgart.de
On Wed, 11 Jul 2012, Tarun Khanna wrote:
hi
I am an undergraduate chemistry student and using espresso 3.1.0 for a short
term
summer project, the system i am currently studying takes around 6-7 hours on a
6 core
machine.I was storing 6000 configurations for offline analysis, so in order to
reduce
the load on the machine i decided to divide the simulation time and store
only 1000
configurations at one time and stopped the simulation there and then next time
I
feed the last configuration stored instead of random initial configuration and
carry
on with the simulation for further 1000 steps so in this way i carried out the
simulation at 6 different times.
But recently when i compared the stored configurations in the above way with the
configuration stored in single run the results were different. so please tell
me where
i am going wrong and also is there any other way in espresso to stop the
simulation
and start from there the next time.
thank you
Tarun