2016-01-29 11:34 GMT-07:00 Chang Woon Jang <[email protected]>: > Dear Christoph, > > I have obtained the forces from csg_boltzmann and csg_resample. When I > run ibi simulation with gromacs, it failed at certain steps even though I > used smaller time step, longer sampling iterations. For example, when I used > 1 fs and 500 ps each step, it failed at 15 steps (step_015). When I used 0.7 > fs and 1 ns each step, it failed at 164 steps (step_164). > > The error is always the same as follow. > > "Fatal error: > 2 of the 900 bonded interactions could not be calculated because some ...." > > Then, I asked GROMACS community. They suggested computing the energy of a > single configuration with "mdrun -s input.tpr -rerun configuration.pdb". > > It produced the following warnings. > > "WARNING: For the 1399 non-zero entries for table 0 in table_b1.xvg the > forces deviate on average 179% from minus the numerical derivative of the > potential" > > When I set all forces "zero", then the warning message is gone. > > Can I just set forces zero or still need to put forces? Are you iterating on the interaction, which uses table_b1.xvg? If not, you can put zeros or edit the table by hand. If you are iterating on that interaction, the table changes every step, and you have to figure out where the numerical derivative comes from, you might need to increase the min of that interactions.
> > Is the meaning of different IBI results the length of convergence or the > final derived potentials? A shorter iteration will lead to a different potential. Due to the fact that it takes 10 times longer to crash, 1fs / 500ps might be a bit too optimistic. Have a look at the *.dist.new files inside step_* to see if the distributions look rough and hence you didn't collect enough statistics. > > Thank you. > > Best regards, > Changwoon Jang > > > > > On Fri, Jan 29, 2016 at 1:15 PM, Christoph Junghans <[email protected]> > wrote: >> >> 2016-01-29 10:56 GMT-07:00 Chang Woon Jang <[email protected]>: >> > Dear Votca Users, >> > >> > I have a question. In the hexane table_b1.xvg in IBI, there are >> > forces >> > in 3rd column, but in the propane table_b1.xvg, there are no forces in >> > 3rd >> > column. >> > >> > Why are there no forces in propane example? How does forces in 3rd >> > column >> > influence the IBI result? >> If all forces are 0, Gromacs will calculate them itself. We prefer >> putting them in the table file to make it more explicit. >> However, we must have forgotten that for this example. >> >> Anyhow, if the spacing is small it should not make a difference >> whether the force are calculated by VOTCA or Gromacs. >> Also Gromacs will warn you if the third column differs from what it >> would calculate internally. >> >> The IBI results certainly depend on the content of table_b1.xvg. >> >> Christoph >> >> > >> > Thank you. >> > >> > Best regards, >> > Changwoon Jang, >> > >> > -- >> > 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. -- 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.
