On Sun, Dec 9, 2018 at 3:13 AM Thomas Meng <thomasmen...@gmail.com> wrote: > > Hi, Christoph. Sorry to bother again! > > At around the 45th IBI iteration using LAMMPS as the sampling program, I came > across a problem saying that "bond atoms are missing". Then I checked the > pair potential I generated, it looks weird and there comes a strange > 'step-shaped' turn at around 3~4 Angstrom. Noticed that the position is near > where the target RDF becomes non-zero, so I wonder why my system didn't > generate a zero RDF before that point. > > Maybe the problem is due to the accumulative effect of the potential before > it? You know, because U_(n+1) = U_(n) + delta_U, so the delta_U before3~4 > Angstrom is always positive in my case and resulted in that problem? What > should I do to solve it?
That usually means something went wrong with the extrapolation on the min side, most likely it wasn't steep enough to prevent particles from coming too close. You will have to play a bit with the r_min for that interaction. Set it so that the rdf is not quite 0 yet, but so that it is very steep. As the average steepness is what determines the extrapolation. Christoph > > 在 2018年12月8日星期六 UTC+8下午9:21:53，Christoph Junghans写道： >> >> On Sat, Dec 8, 2018 at 1:05 AM Thomas Meng <thomas...@gmail.com> wrote: >> > >> > Thanks, I think I understood what you said about the topol.xml's bonded >> > definition thing. In order to perform the ibi >> > calculation, it seems that I need to write a new topol.xml for my >> > coarse-grained system. Besides, I could also use >> > a lammps data file as a topol file as you mentioned before with a little >> > modification to the setting.xml file. However, the >> > use of lammps data as a topol may not be appropriate for ibi calculation. >> > >> > I tested my polyether system using both the CG configuration' data and a >> > house-written topol.xml as the topol file to >> > realize the ibi procedure.Then I compared the resulted RDF generated by >> > these calculations after only one iteration, >> > and they are just different (I attached the figures). It could be >> > obviously observed that the peak of RDF is a bit higher >> > than the target one, however that is very similiar to the RDF without >> > considering those bond interactions which I calculated >> > before. So, I guess that as a topo file, the LAMMPS data didn't consider >> > the bonded interaction correctly. >> Yes, I think you are correct, in the RDF from the data files no >> exclusions seemed to generated, "csg_dump --excl" should confirm that >> as well. >> So I guess we still need to improve on the data reader a bit. >> >> Christoph >> > >> > 在 2018年12月7日星期五 UTC+8下午8:49:08，Christoph Junghans写道： >> >> >> >> On Thu, Dec 6, 2018 at 7:29 PM <thomas...@gmail.com> wrote: >> >> > >> >> > 在 2018年12月7日星期五 UTC+8上午6:07:29，Christoph Junghans写道： >> >> >> >> >> >> On Thu, Dec 6, 2018 at 9:29 AM <thomas...@gmail.com> wrote: >> >> >> > >> >> >> > Hi, >> >> >> > >> >> >> > I'm using VOTCA to coarse-grain a relatively simple polyether >> >> >> > system, [C-C-O-C-C]n, and a super atom just represents a single >> >> >> > monomer. After performing an all-atom simulation in LAMMPS, I got a >> >> >> > trajectory file (trj_tail20.dump) and a final data file >> >> >> > (nvt_300K_final.data). I wrote my own map and topol .xml files, and >> >> >> > then use them to calculate the RDF of the system. >> >> >> > >> >> >> > What got me frustrated is that, the RDF I obtained using csg_stat >> >> >> > (figure AA.png) seems to be questionable. Because if I use LAMMPS to >> >> >> > generate the RDF of the central O atom using the final data file, I >> >> >> > got a different result (AA-LAMMPS.png). And the most strange thing >> >> >> > is that if I use OVITO to analyze the CG configuration I converted >> >> >> > using the csg_map command, I got a similar result compared to that >> >> >> > of lammps simulation (AA-OVITO.png). >> >> >> > >> >> >> > Because the RDF is a quite important intermediate process file, so I >> >> >> > wonder what resulted in the above difference? And could I trust the >> >> >> > one generated through VOTCA? >> >> >> >> >> >> csg_stat has been tested over years, so I don't think there is a >> >> >> problem in the code. >> >> >> But I saw your topol.xml has a <bond> section, which in VOTCA >> >> >> generates a bunch of RDF exclusions, i.e. pairs in a bond aren't >> >> >> considered for the RDF calculation. You can see that running: >> >> >> $ csg_dump --top topol.xml --excl >> >> >> >> >> >> So if you don't want that just delete that section. >> >> >> >> >> > Yes, I've just tested it, the difference surely comes from the bond >> >> > definition. However, it's not the >> >> > bond info in the topol.xml file, but the </cg_bonded> section in the >> >> > mapping 'cg.xml' file. If I just >> >> > delete that section, then I will get the same RDF as the one generated >> >> > from LAMMPS (I attached >> >> > the figure). >> >> You are right, it is the <cg_bonded> block as you are using a mapping >> >> file, which then sets up bonds (with exclusions) in the mapped system. >> >> >> >> However, on a second thought, I don't think the bonds should be >> >> included in RDF unless you are planning to use the non-bonded >> >> interactions also between cg beads in the same molecule. >> >> >> >> > >> >> > Thanks. And now I will think about which one is suitable for my case. >> >> > >> >> >> Also in the development version of VOTCA you can use LAMMPS' data >> >> >> files as topology input directly. >> >> >> >> >> > But if I use the LAMMPS data as a top file, then there will be many >> >> > unkown molecule problems (it >> >> > may comes from LAMMPS use numbers to represent atoms and molecules), so >> >> > I would rather write >> >> > a topol.xml myself. However, from my previous test, a data could be >> >> > used as a trjactory file, though >> >> > it only contains one frame, may generate a rough RDF curve. >> >> Please open an issue on github for each of the "unkown molecule >> >> problems", maybe we can help fix the code or your input file. >> >> >> >> Christoph >> >> > >> >> > Anyway, thank you very much for your reply. :) >> >> > >> >> > Meng >> >> > >> >> >> Christoph >> >> >> >> >> >> > >> >> >> > Meng >> >> >> > >> >> >> > P.S. >> >> >> > The command I used to generate RDF is "csg_stat --top topol.xml >> >> >> > --trj trj.dump --cg cg.xml --options settings.xml" >> >> >> > >> >> >> > -- >> >> >> > 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 >> >> > >> >> > -- >> >> > 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 >> > >> > -- >> > 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 > > -- > 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+unsubscr...@googlegroups.com. > To post to this group, send email to votca@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 -- 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+unsubscr...@googlegroups.com. To post to this group, send email to votca@googlegroups.com. Visit this group at https://groups.google.com/group/votca. For more options, visit https://groups.google.com/d/optout.
