[gmx-users] pull code in 5.1.2

Wed Mar 2 21:48:20 CET 2016

Hi,

Per the new
http://manual.gromacs.org/documentation/5.1/ReleaseNotes/new-features.html#changed-the-way-pull-directions-are-selected,
the
old syntax didn't permit you to have multiple pull coordinates that
differed in either type or geometry. We thought that was an unwarranted
restriction. It could have been a bit better if we'd continued to support
the old syntax, but that just defers the pain for a year or so. :-)

Mark

On Wed, Mar 2, 2016 at 9:23 PM Alex <nedomacho at gmail.com> wrote:

> Mark, thanks for responding. It seems like this is the second syntax
> change for COM pulling since v. 4.6.x
> If you don't mind my asking, why is this happening, esp. given no
> backward compatibility? By comparing e.g. 5.0 and 5.1, I don't readily
> see the rationale...
>
> Peter, your suggested syntax works out of the box, thank you.
>
> Alex
>
> On 3/2/2016 2:16 AM, Mark Abraham wrote:
> > Hi,
> >
> > Yes, Google continues to find both the old and new documentation, while I
> > haven't found the time to migrate everything else to the new layout. The
> > documentation you found is marked as being for 5.0...
> >
> > In general, seek documentation at
> http://manual.gromacs.org/documentation/.
> > I have updated the 5.1 release notes to be more specific on this
> particular
> > topic.
> >
> > Mark
> >
> >
> >
> > On Wed, Mar 2, 2016 at 9:57 AM Alex <nedomacho at gmail.com> wrote:
> >
> >> Okay, I am officially the Jon Snow of COM pulling under 5.1.2. :) At the
> >> same time, neither the new manual, nor the google-preferred page
> >> (http://manual.gromacs.org/online/mdp_opt.html#pull) have the info you
> >> just linked to.
> >>
> >> Thanks a lot, let me try this.
> >>
> >> Alex
> >>
> >>
> >> On 3/2/2016 1:44 AM, Peter Kroon wrote:
> >>> Hi,
> >>> According to the manual
> >>> (
> >>
> http://manual.gromacs.org/documentation/5.1/user-guide/mdp-options.html#com-pulling
> >> )
> >>> it should be something like this:
> >>>
> >>> pull = yes
> >>> pull-nstxout        = 1000
> >>> pull-nstfout        = 1000
> >>> pull_ngroups = 1
> >>> pull_group1_name = JON_SNOW_KNOWS_NOTHING
> >>> pull_ncoords = 1
> >>> pull-coord1-type = umbrella
> >>> pull_coord1_geometry = direction-periodic
> >>> pull_coord1_groups = 0 1 ; This means you are using an absolute
> >>> reference position (group 0).
> >>> pull_coord1_origin = 0 0 0 ; This is also the default value
> >>> pull_coord1_vec = 0 0 -1
> >>> pull_coord1_rate    = 5.0e-5
> >>> pull_coord1_k       = 1000
> >>> pull_coord1_start = yes
> >>>
> >>> Read the manual carefully, and make sure my suggestions are correct for
> >>> your case.
> >>> It might be nice to have a section in the manual (or a webpage) that
> >>> tells you how to translate between mdp versions.
> >>>
> >>> Peter
> >>>
> >>> On 02/03/16 07:48, Alex wrote:
> >>>> Hi all,
> >>>>
> >>>> In the Game of Thrones, as soon as you learn characters' names, they
> >>>> get brutally murdered. As soon as I got a good command of the pull
> >>>> code, an mdp working perfectly in 5.0.4 crapped its pants in 5.1.2
> >>>> (quoted below). The latest manual no longer even has a syntax section
> >>>> for pull code. 5.1.2 wasn't exactly my idea, we just got a new GPU
> >>>> node and the admin installed it. Roll back to older version?
> >>>>
> >>>> Thanks,
> >>>>
> >>>> Alex
> >>>>
> >>>> pull code section from a working mdp:
> >>>>
> >>>> ***
> >>>> pull                = umbrella
> >>>> pull_geometry       = direction-periodic
> >>>> pull_coord1_vec     = 0.0 0.0 -1.0
> >>>> pull_start          = yes
> >>>> pull_ngroups        = 1
> >>>> pull_ncoords        = 1
> >>>> pull_coord1_groups  = 0 1
> >>>> pull_group1_name    = JON_SNOWS_DEAD_BODY
> >>>> pull_coord1_rate    = 5.0e-5
> >>>> pull_coord1_k       = 1000
> >>>> pull-nstxout        = 1000
> >>>> pull-nstfout        = 1000
> >>>> ***
> >>>>
> >>>> Complaints below. Curiously, they all point to the same line, and that
> >>>> is a line nowhere near the pull code.
> >>>>
> >>>> WARNING 1 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_geometry' in parameter file
> >>>>
> >>>> WARNING 2 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_coord1_vec' in parameter file
> >>>>
> >>>> WARNING 3 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_start' in parameter file
> >>>>
> >>>> WARNING 4 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_ngroups' in parameter file
> >>>>
> >>>> WARNING 5 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_ncoords' in parameter file
> >>>>
> >>>> WARNING 6 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_coord1_groups' in parameter file
> >>>>
> >>>> WARNING 7 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_group1_name' in parameter file
> >>>>
> >>>> WARNING 8 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_coord1_rate' in parameter file
> >>>>
> >>>> WARNING 9 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull_coord1_k' in parameter file
> >>>>
> >>>> WARNING 10 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull-nstxout' in parameter file
> >>>>
> >>>> WARNING 11 [file pull_300ns.mdp, line 51]:
> >>>>     Unknown left-hand 'pull-nstfout' in parameter file
> >>>>
> >>>
> >>>
> >> --
> >> Gromacs Users mailing list
> >>
> >> * Please search the archive at
> >> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> >> posting!
> >>
> >> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >>
> >> * For (un)subscribe requests visit
> >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> >> send a mail to gmx-users-request at gromacs.org.
> >>
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>