Hi Mohammad,

The extrapolated potential for r < rmin must be repulsive always, otherwise
CG-MD simulations will crash.

Both the linear and exponential extrapolations (given by following
expressions) use the slope of the potential at rmin from relative entropy
update.

'm = dy/dx
- constant:  ''\$y = y0\$''
- linear:  ''\$y = ax + b\\;\\;b = - m*x_0 + y_0\;\;a = m\$''
- exponential: ''\$y = a*\\\\exp(b*x)\\;\\;a = y0*\\\\exp(-m*x0/y0)\\;\\;b
= m/y_0\$''

However, exponential also uses the value of the potential at rmin. Hence,
at some relative entropy update, if the value of the potential at rmin
becomes negative, even though the slope is repulsive, then the extrapolated
potential may become attractive and the following CG-MD run may crash.
Therefore, the linear extrapolation is preferred.

Exponential extrapolation generally results in the more repulsive core than
the linear extrapolation. However, with careful choice of rmin based on the
target RDF, linear extrapolation also results in sufficiently repulsive
core. Hence, there may not  be much difference between the CG-MD
simulations with exponentially extrapolated potentials and that with the
linearly extrapolated potentials.

I hope above explanation answers your questions.

Cheers,
Sikandar

On Tue, Jul 15, 2014 at 7:06 PM, Hossein Motevaselian <
[email protected]> wrote:

> Hi,
>
> I was looking at the relative entropy code and find out the current
> version only linearly extrapolates the knot values less than rmin.
> Is there any reason not to extrapolate exponentially? And does it make a
> difference?
>
> Sincerely,
> Mohammad
>
> --
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at http://groups.google.com/group/votca.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"votca" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/votca.
For more options, visit https://groups.google.com/d/optout.

Reply via email to