Re: CCL:MD at constant temperature

From: Kenneth Geisshirt <kenneth %! at !% geisshirt.dk>
Subject: Re: CCL:MD at constant temperature
Date: Thu, 20 Dec 2001 09:03:12 +0100 (CET)

On Thu, 20 Dec 2001, [gb2312] yong pei wrote:
 > The gear predict-corrector method was used .
 Why don't you use a symplectic algorithm? A symplectic algorithm is
 typical numerically more stable and predictor-corrector methods.
 > Then, how can I determine the initial
 > values of s.derivate of s in the formular?
 I would begin with s = 0 and take a few steps with a simple integrator.
 Btw, you're more than welcome to download my old MD software and take a
 look at my Nose-Hoover implementation or read my Ph.D. thesis (there is a
 chapter on numerical algorithms):
 http://kenneth.geisshirt.dk/software/mdreac.tar.gz
 http://kenneth.geisshirt.dk/publications/phd-thesis.ps.gz
 Kneth
 --
 Kenneth Geisshirt, M.Sc., Ph.D.         http://kenneth.geisshirt.dk
 Grøndals Parkvej 2A, 3. sal                    kenneth %! at !%
 geisshirt.dk
 DK-2720 Vanløse                                     +45 38 87 78 38