Re: Namd, lammps comparing energies

From: Giacomo Fiorin (
Date: Fri Aug 10 2012 - 13:20:21 CDT

Hello Jackie, to what precision are the energies equal? If you used a PDB
input or any other formats with few digits of precision, energies will of
course be different just because of the roundoff.

It also goes without saying that the two configurations must be identical
(the same topology file is not enough), and you should take care of all
options that different default values in the two programs.

Also, you must compare energies *only* at the step number 0, or at the end
of a minimization. When you're running at finite temperature and pressure
(implied by your error message), you are bound to see slightly different
things, starting from step 1.


On Fri, Aug 10, 2012 at 1:44 PM, Jacqueline Schmidt <> wrote:

> Dear NAMD users,
> I have a question regarding how the energies are compared in lammps and
> namd. I have set the parm (for namd) compatible with my lammps data file
> but I get different energies (even for bond and angle the energy that I get
> in namd is much lower than lammps). The structures are exactly the same and
> I want to make sure I get the same initial energies (at least close) in the
> two packages (I tried this with a simpler system containing only one
> molecule but namd gives an error of " Periodic cell has become too small
> for original patch grid" which I couldn't solve by increasing the margin or
> use "useFlexibleCell no" command). Can anyone make a comment please?
> Thanks
> Jackie

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:22:22 CST