Hi Marc The advantage of Cubic B-splines functional form is that it is very flexible and it can capture any shape provided that sufficiently large number of knot values are used.
The knot values determine the grid-size for B-splines, i.e., dr = Rcut/(nknots-3). Smaller the grid size more flexible is the B-spline. But decreasing grid size, i.e. more knot values, increases computational time of RE. So you have to estimate the minimum number of knot values that will capture the CG potential shape. In the case of water, CG potential is known to have two-minima core-softened shape. Hence to capture such a shape we used 48 knot values to model B-Splines. Since, we use cut-off of 0.9 nm, the grid size for B-spline is 0.02 nm (cut-off/(nknots-3)). I hope this helps. Regards Sikandar On Mon, Apr 14, 2014 at 12:50 PM, Marc Segovia <[email protected]>wrote: > Hi all > > I am running RE tutorial in develpment version, but a question arise. > When is estimated the potential using cubic splines, how do you > specifed the number of knots ? in the tutorial they are 48. > > <!-- function form for the potential (lj126 or ljg or cbspl)--> > <function>cbspl</function> > <cbspl> > <!-- no. of knots for cbspl function --> > <nknots>48</nknots> > > Thanking in advance your help > > Best Regards > > Marc > > Dr. Marc Segovia > CCBG-DETEMA > Facultad de Quimica > Universidad de la Republica > 11800 Montevideo > Uruguay > > > > On Fri, Nov 1, 2013 at 11:57 PM, Sikandar Mashayak > <[email protected]>wrote: > >> Hi Marc >> >> The option "nknots_cutoff" was earlier used in the settings.xml file to >> specify number of cubic B-spline knots near the cut off which should be >> forced to value 0.0 to ensure that CG potential smoothly goes to zero near >> the cut-off. >> >> Later, I decided to fix this nknots_cutoff = 4 because I think fixing 4 >> knot values near the cut-off is sufficient in most of the cases and user >> should not worry about this number. Hence, instead of taking this number as >> an input from user, it is fixed in the source code. >> >> You should not get the error "property not found: nknots_cutoff" if you >> are using latest source code of votca-csg (i.e., revision > >> 681d33b7942b<http://code.google.com/p/votca/source/detail?r=681d33b7942bba2d3d213f4e321b17e0c6a00676&repo=csg>). >> >> >> So could you please check if you are correctly downloading the latest >> version and installing it. Because, I dont get the same error, when I test >> the water spce relative entropy tutorial with the latest development >> version of csg source code. >> >> You can check if you are using the latest version of votca-csg by running >> "csg_inverse -h" it will print out the version of votca-csg that you are >> using. >> >> Let us know if you face issues with updating and installing the latest >> development version of votca-csg. >> >> Thanks >> Sikandar >> >> >> >> On Fri, Nov 1, 2013 at 6:03 PM, Marc Segovia <[email protected]>wrote: >> >>> Good Night Everyone >>> >>> I am trying to run re tutorial with spce, but an error appear about >>> nknots_cutoff property, I update my dev version without sucess, any >>> suggestion ? >>> >>> Wishing a very good weekend and thanking in advance . >>> >>> Marc >>> >>> Dr. Marc Segovia >>> CCBG >>> Facultad de Quimica >>> Universidad de la Republica >>> 11800 Montevideo >>> Uruguay >>> >>> ~/votca1.3/csg-tutorials.votca-695b567ead5c/spce/re/step_000 >>> Generating potential tables from the initial parameters >>> Running critical command 'csg_reupdate --gentable true --param-in-ext >>> param.new --options >>> /home/marc/votca1.3/csg-tutorials.votca-695b567ead5c/spce/re/settings.xml' >>> an error occurred: >>> property not found: nknots_cutoff >>> >>> Callstack: >>> /home/marc/votca1.3/share/votca/scripts/inverse/inverse.sh - linenumber >>> 166 >>> do_external - linenumber 171 in >>> /home/marc/votca1.3/share/votca/scripts/inverse/functions_common.sh >>> /home/marc/votca1.3/share/votca/scripts/inverse/prepare_re.sh - >>> linenumber 36 >>> critical - linenumber 4 (see 'csg_call --cat function >>> critical') >>> die - linenumber 2 (see 'csg_call --cat function die') >>> >>> ################################################################################################################################################################ >>> # >>> # >>> # >>> ERROR: >>> # >>> # critical: 'csg_reupdate --gentable true --param-in-ext param.new >>> --options >>> /home/marc/votca1.3/csg-tutorials.votca-695b567ead5c/spce/re/settings.xml' >>> failed # >>> # For details see the logfile >>> /home/marc/votca1.3/csg-tutorials.votca-695b567ead5c/spce/re/inverse.log >>> # >>> # >>> # >>> >>> ################################################################################################################################################################ >>> die: (called from 318) CSG_MASTER_PID is 31799 >>> die: pids to kill: 31799 318 >>> >>> -- >>> 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 http://groups.google.com/group/votca. >>> For more options, visit https://groups.google.com/groups/opt_out. >>> >> >> -- >> 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 http://groups.google.com/group/votca. >> For more options, visit https://groups.google.com/groups/opt_out. >> > > -- > 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 http://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 http://groups.google.com/group/votca. For more options, visit https://groups.google.com/d/optout.
