dumping out potential energy for alternate potentials

Hi Daniel,
It would be very helpful if you can post the files for doing the weighted pair/hybrid/overlay evaluation. I’m interested!

Ok, here it comes. Steve mentioned the pair compute, which I was not
aware of, so my solution was to write a small dummy compute which the
total energy values get saved into (and which performes the mpi
reduction). I should probably rewrite my stuff to use the existing
compute.
The weights for the pair sub styles are stored in a new member added
to the pair base class and are set using a pair_modify command (option
"weights" followed by n float values in the range [0..1], where n is
the number of substyles in the pair/hybrid/weighted style).

I attached an example input that should illustrate the syntax. The
store/pe compute is automatically generated with the name 'hype' by
the pair style. You can access the values as c_hype[1] - c_hype[n]

Best,
Daniel

P.S.: Let me know what you need this for.

pair.patch (1.77 KB)

compute_storepe.cpp (2.2 KB)

compute_storepe.h (726 Bytes)

pair_hybrid_weighted.cpp (6.34 KB)

pair_hybrid_weighted.h (1.14 KB)

in.hyweigh (740 Bytes)