Bond Potential Extrapolation

[Joshua Moore]

I am having issues with the bond potential table written during ibi.  If I cut it only in the region that should be fit (where the distribution is nonzero), it does not extend far enough for LAMMPS, but if I extend this region in the bonded section, the extrapolation is not steep enough.  Please see the attachment.  LAMMPS recently changed so if a bond length is outside the table it gives an error.  Previously it would just use the force at the end points.

 

Is there an option to set only a specific range, but then extend this when writing the table?

Do we need to set the upper and lower end points like the angle potential in potential_to_lammps.sh?  It looks like there is an option for Gromacs but not LAMMPS?

  <bonded>
    <!-- name of the interaction -->
    <name>A_A_bond</name>
    <min>2.0</min>
    <max>8.0</max>
    <step>0.01</step>
    <inverse>
      <!-- target distribution -->
      <target>A_A_bond.dist.tgt</target>
      <do_potential>0</do_potential>
      <lammps>
        <table>table_b1.pot</table>
      </lammps>
    </inverse>
  </bonded>

Thanks.

Josh

[Christoph Junghans]

2017-03-10 0:05 GMT-07:00 Joshua Moore <jdm...@ncsu.edu>:
> I am having issues with the bond potential table written during ibi. If I
> cut it only in the region that should be fit (where the distribution is
> nonzero), it does not extend far enough for LAMMPS, but if I extend this
> region in the bonded section, the extrapolation is not steep enough. Please
> see the attachment. LAMMPS recently changed so if a bond length is outside
> the table it gives an error. Previously it would just use the force at the
> end points.
>
> Is there an option to set only a specific range, but then extend this when
> writing the table?
>
> Do we need to set the upper and lower end points like the angle potential in
> potential_to_lammps.sh? It looks like there is an option for Gromacs but
> not LAMMPS?

Apparently, there was only a <table_begin> option for lammps, but no
<table_end> option (like gromacs has), so I added that:
<https://github.com/votca/csg/commit/c228c63844efe8789cd0242f1f592b0519c3e701>

Christoph

>
> <bonded>
> <!-- name of the interaction -->
> <name>A_A_bond</name>
> <min>2.0</min>
> <max>8.0</max>
> <step>0.01</step>
> <inverse>
> <!-- target distribution -->
> <target>A_A_bond.dist.tgt</target>
> <do_potential>0</do_potential>
> <lammps>
> <table>table_b1.pot</table>
> </lammps>
> </inverse>
> </bonded>
>
> Thanks.
>
> Josh
>

> --
> 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 votca+un...@googlegroups.com.
> To post to this group, send email to vo...@googlegroups.com.
> Visit this group at https://groups.google.com/group/votca.
> For more options, visit https://groups.google.com/d/optout.



--
Christoph Junghans
Web: http://www.compphys.de

[Joshua Moore]

Thanks!

I had just added that and was going to do a pull request but you beat me to it.

That works much better now with that option.

I was still having an issue though with the extrapolation, so I changed the bond, angle and torsion extrapolation to the sasha function in potential_extrapolate.sh. and I'm having a lot less issues now.

Is it possible to control this option for the extrapolation option from within the xml?  It looked like I could only change it within potential_extrapolate.sh.

Thanks.

Josh

[Christoph Junghans]

2017-03-10 12:01 GMT-07:00 Joshua Moore <jdm...@ncsu.edu>:
> Thanks!
>
> I had just added that and was going to do a pull request but you beat me to
> it.
>
> That works much better now with that option.
>
> I was still having an issue though with the extrapolation, so I changed the
> bond, angle and torsion extrapolation to the sasha function in
> potential_extrapolate.sh. and I'm having a lot less issues now.
>
> Is it possible to control this option for the extrapolation option from
> within the xml? It looked like I could only change it within
> potential_extrapolate.sh.

Does this work for you: <https://github.com/votca/csg/pull/203> ?