BoltzmaNN: Predict pair potential using neural networks
Input g(r)
Input density
Output U(r)
g(r)
U(r)
Correct for hard-core contribution
Choose a model:
Model v.1.0
Model v.1.1
Model v.1.2
Calculate U(r)
Load example
Usage
This tool computes the effective pair potential
U
(
r
) from the radial pair distribution function
g
(
r
) using neural networks (see
https://gitlab.rlp.net/anikouba/boltzmann
and
https://arxiv.org/abs/1908.02448
for details).
Input:
g
(
r
) values (200 values, sampled between
r
=0 and 20) and number density in units of 1/sigma^3
Output:
U
(
r
) values (50 values, sampled between
r
=0 and 5)
Tools developed by
Fabian Berressem
and
Arash Nikoubashman