Hi, Currently RE implementation does not support constraint optimization. So only way to ensure physically meaningful parameters (C12, C6 or sigma epsilon) is to provide good initial guess, check other MD simulation parameters such as cut-off, dielectric permittivity in case of charges, and using small enough relaxation/mixing parameter.
Best, Sikandar On Sat, Apr 9, 2016 at 8:31 PM, Christoph Junghans <[email protected]> wrote: > 2016-04-09 16:38 GMT-06:00 Alireza Moradzadeh <[email protected]>: > > I looked at simplex tutorials. I don,t now why we need n+1 line, > n+1 sets is a requirement of the simplex algorigtm, see our recent paper: > <http://dx.doi.org/10.1371/journal.pone.0131754> > or > <https://en.wikipedia.org/wiki/Nelder%E2%80%93Mead_method> > > furthermore, when i tried to run my settings.xml based on the a > combination rule I got the error mentioning there is no parameters for > > interaction between two beads. I am trying simplex on a four bead model > so I don't have enough parameters to allocate for all of the > > interaction. > Like I said in my last email you need at least one parameter per > interaction, you have 3 interactions and 4 parameters so you should be > fine. However looking at the code, I think it could be rewritten to > list this constraint. > > Christoph > > > >> On Apr 9, 2016, at 2:49 PM, Christoph Junghans <[email protected]> > wrote: > >> > >> 2016-04-08 22:26 GMT-06:00 Alireza Moradzadeh <[email protected]>: > >>> By the way I would be glad if you could provide more information on > how to > >>> choose parameters initials > >>> I could not find so much information on how I should choose those > value or > >>> even its organization > >> Have a look at the methanol-water simplex tutorial there is a > >> simplex.in file for each interaction. > >> The values in there are given in the same order as in the xml file. > >> You will need n+1 sets to start with (lines in simplex.in) These sets > >> can be generated by randomly perturbing some values from a single set, > >> which you got from e.g. fitting an IBI potential. > >> > >> Christoph > >> > >>> > >>> Thanks, > >>> Alireza > >>> > >>> > >>>> On Friday, April 8, 2016 at 10:21:23 PM UTC-5, Alireza Moradzadeh > wrote: > >>>> > >>>> Thank you for your answer. > >>>> > >>>> I am also wondering if there is a way to keep eps and sigma positive > >>>> during simulation time. Because the optimization method is blind to > physical > >>>> meaning of eps and sigma I got negative value using other methods in > votca > >>>> package. > >>>> Is it possible to enforce this condition using this method or RE. > Also is > >>>> there anyway in votca package to treat charge of beads as a variable > during > >>>> simulation so I could control sigma and epsilon by variable charge. > >>>> > >>>> Thanks, > >>>> Alireza > >>>> > >>>> > >>>>> On Friday, April 8, 2016 at 4:04:48 PM UTC-5, Christoph Junghans > wrote: > >>>>> > >>>>> 2016-04-07 19:10 GMT-06:00 Alireza Moradzadeh <[email protected]>: > >>>>>> Dear VOTCA Users, > >>>>>> > >>>>>> > >>>>>> For a system composed of two beads , I want to optimize pressure and > >>>>>> interaction based on RDF but I want to stick to Lorentz-Berthelot > >>>>>> combination rule. I want to know if I define my function as follows > >>>>>> would it > >>>>>> be right or not? and should I define pressure as target in each > >>>>>> interaction. > >>>>> Parameters are always global so, you should not list any parameter > >>>>> twice, but you need to list at least one parameter for each > >>>>> interaction. > >>>>> e.g. list eps1 sig1 for CG1-CG1, eps2 for CG1-CG2 and sig2 for > CG2-CG2 > >>>>>> > >>>>>> CG1-CG1 > >>>>>> target rdf presssure > >>>>>> function 4*eps1*[(sig1/x)**12-(sig1/x)**6] > >>>>>> parmeters eps1 sig1 > >>>>>> > >>>>>> CG1-CG2 > >>>>>> target rdf pressure > >>>>>> function > 4*(eps1*eps2)**2*[((sig1+sig2)/2/x)**12-(sig1+sig2)/2/x)**6] > >>>>> This should be square root and not Power 2. > >>>>> > >>>>> Christoph > >>>>>> parameters eps1 sig1 eps2 sig2 > >>>>>> > >>>>>> > >>>>>> CG2-CG2 > >>>>>> target rdf presssure > >>>>>> function 4*eps2*[(sig2/x)**12-(sig2/x)**6] > >>>>>> parmeters eps2 sig2 > >>>>>> > >>>>>> Thanks, > >>>>>> > >>>>>> -- > >>>>>> 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 https://groups.google.com/group/votca. > >>>>>> For more options, visit https://groups.google.com/d/optout. > >>>>> > >>>>> > >>>>> > >>>>> -- > >>>>> Christoph Junghans > >>>>> Web: http://www.compphys.de > >>> > >>> -- > >>> 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 https://groups.google.com/group/votca. > >>> For more options, visit https://groups.google.com/d/optout. > >> > >> > >> > >> -- > >> Christoph Junghans > >> Web: http://www.compphys.de > >> > >> -- > >> You received this message because you are subscribed to a topic in the > Google Groups "votca" group. > >> To unsubscribe from this topic, visit > https://groups.google.com/d/topic/votca/ISshiaVA4mc/unsubscribe. > >> To unsubscribe from this group and all its topics, send an email to > [email protected]. > >> To post to this group, send email to [email protected]. > >> Visit this group at https://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 https://groups.google.com/group/votca. > > For more options, visit https://groups.google.com/d/optout. > > > > -- > Christoph Junghans > Web: http://www.compphys.de > > -- > 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 https://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 https://groups.google.com/group/votca. For more options, visit https://groups.google.com/d/optout.
