Hey Christoph You are right, from looking in the literature it does seem complicated.. and my idea of doing some interpolation is not very rightful because by doing that I would be introducing "forces" at given pairwise distances which do not come from predictions from the force matching method. I mean, if the whole purpose is to find a potential that reproduces the forces, I think I would jeopardize the whole thing to one extent or another if my superatoms end up being, even if for a tiny moment, in the artificial repulsive regions I was planning to introduce.
I saw some people who do FM at high temperature in order to get the separate peaks to merge. Seems to have worked for me for the peaks sitting at > 10 angs. And since my first and second peaks in the RDF I showed comes from 1-2 and 1-3 neighbors, I could assign them a separate bonded potential instead. I still have some other problems in the potential I am obtaining but I will figure out :)) Thanks! Best, Cecilia Em sábado, 3 de junho de 2023 às 16:20:38 UTC+2, Christoph Junghans escreveu: > On Wed, May 31, 2023 at 4:27 AM Cecília Álvares > <[email protected]> wrote: > > > > Hmm.. maybe this is actually a very stupid question, because, given that > the potential is built by using splines that hold within a small > subinterval defined in <step>, I suppose I could get the same result by > splitting the ranges of non-zero RDF in different interactions. (it should > give the same results, right?) > > The reason why we haven't implemented support for non-continuous > domains is mainly that it is complicated to get all the cases right > and so we thought one could just do a piecewise ibi and then combine > it. > > Christoph > > > > > Em quarta-feira, 31 de maio de 2023 às 12:07:35 UTC+2, Cecília Álvares > escreveu: > >> > >> (I mean, in this case, I could get "chops" of how the potential would > look like in different intervals (corresponding to g(r) != 0) and I later > can connected the different portions by doing some sort of interpolation > myself.) > >> > >> Em quarta-feira, 31 de maio de 2023 às 11:54:18 UTC+2, Cecília Álvares > escreveu: > >>> > >>> Hello everyone, > >>> > >>> I was wondering if there is a way to define a "non-continuous" domain > interval when setting up the settings.xml file to do force-matching of > non-bonded potentials (for example) within VOTCA. I mean, in my case, since > I am deriving potentials for a xtalline solid, I have peaks in the RDF that > go up and down back to 0 (e.g. below), which would give rise to non-sampled > intervals for the potential if I was to use a continuous interval for rij > going from the first peak to the last. > >>> > >>> > > -- > > Join us on Slack: https://join.slack.com/t/votca/signup > > --- > > 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 view this discussion on the web visit > https://groups.google.com/d/msgid/votca/c201cbeb-641d-4a7c-b043-d55101cfcb78n%40googlegroups.com > . > > > > -- > Christoph Junghans > Web: http://www.compphys.de > -- Join us on Slack: https://join.slack.com/t/votca/signup --- 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 view this discussion on the web visit https://groups.google.com/d/msgid/votca/c36825e2-0b1e-4864-94f2-897d61d78dden%40googlegroups.com.
