Project

General

Profile

Bug #385

integrator = steep; nsteps = 0 runs an infinite number of steps

Added by Chris Neale almost 10 years ago. Updated over 9 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Erik Lindahl
Category:
mdrun
Target version:
Affected version - extra info:
Affected version:
Difficulty:
uncategorized
Close

Description

Perhaps this is a feature, rather than a bug, and I can see why one would want to do this some times. However, the behaviour does not match up with integrator = md; nsteps = 0, for which a zero-step run is actually completed.

In any event, I could not find any documentation for this in the online or pdf manual.

History

#1 Updated by Erik Lindahl almost 10 years ago

Hi Chris,

I do indeed think it's a "feature"; for a while it was present for integrator=md too, but there I reverted it since I frequently use nsteps=0 for testing energies/forces.

David/Berk: Unless you protest, I think we should change it so "0" means a single potential evaluation,but no steps, and use e.g. "-1" to mean "run forever".

Cheers,

Erik

#2 Updated by Chris Neale almost 10 years ago

I think that is a great idea. It would be even better if this became generally true such that md, bd, and sd also run forever if nsteps=-1 (not currently possible as far as I know) and simply do the energy evaluation with no steps if nsteps=0 (currently true).

#3 Updated by Berk Hess almost 10 years ago

I wonder if all the minimizers currently behave the same with nsteps=0 or 1.
But I agree with implemented the ideas for nsteps=0 and nsteps=1.
Especially with the current -maxh feature is it useful to have an
unlimited amount of steps.

Berk

#4 Updated by David van der Spoel over 9 years ago

Has Erik's idea been implemented?

#5 Updated by Berk Hess over 9 years ago

I implemented nsteps=0 is 0 steps, nsteps=-1 is unlimited.
-maxh does not work for the minimizers, but I guess that in most
cases you will reach machine precision limits quite soon.

Berk

Also available in: Atom PDF