RE: wrong simulation times or misleading time counting?

From: JC Gumbart (gumbart_at_ks.uiuc.edu)
Date: Wed Jan 18 2006 - 23:27:28 CST

I am almost sure that namdplot never considers the timestep and I imagine
mdenergy doesn't either. This would explain why it's reading 5 million
steps as 5 ns. I think you are actually okay.

-----Original Message-----
From: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] On Behalf
Of Carlos Simoes
Sent: Wednesday, January 18, 2006 5:49 PM
To: namd-l_at_ks.uiuc.edu
Subject: namd-l: wrong simulation times or misleading time counting?

Dear all,

    I started using NAMD a couple of months ago and I've been running
some simulations with a 2 fs iteration timestep. More recently, I
noticed that both namdplot and mdenergy analysis over the entire
simulations show them to have only 5 ns instead of 10 ns (as I would
expect), and I'm trying to figure out why. NAMD User's Guide states
that the total amount of simulation time is given by "numsteps x
timestep". Hence, if I'm carrying out one run for 5 000 000 steps,
with a timestep of 2 fs, I shall expect to get a 10 000 000 fs (10 ns)
simulation. Am I right? Moreover, I suppose it won't make any
difference whether I use the "numsteps" flag or the "run" command
(I've been employing the latter). Obviously, I might have missed a
flag, while using the mdenergy script, but, strangely, it correctly
recognizes the 2 fs timestep (when it starts reading the dcd file).

    Thanks for any tip you can come up with, in advance.

Carlos Simoes

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:41:32 CST