Re: colvars: Error: Feature unavailable:compute total collective force only from one group center

From: Giacomo Fiorin (giacomo.fiorin_at_gmail.com)
Date: Thu Feb 23 2017 - 21:50:08 CST

Hi Brian, the new harmonicWalls bias indeed replaces the options lowerWall
and upperWall, but those will still work.

In general you don't need to use the new syntax if you're continuing an
older simulation, Colvars is generally backward-compatible (with a single
exception: https://colvars.github.io/totalforce.html).

You are clearly not using the default settings if you are defining
lowerWalls explicitly, so please do not state that the default settings are
not correct.

Regarding the error, the manual does not need to cover what is explained by
the error message. If you apply only a lower wall to an angle, sooner or
later the angle will increase beyond +180° and will appear to the left of
the lower wall, with an exceedingly high energy.

Giacomo

On Thu, Feb 23, 2017 at 3:58 PM, Bennion, Brian <bennion1_at_llnl.gov> wrote:

> Ok. So it seems that all lower and upperWall forces as well as the
> constants have been deprecated and replaced with harmonicWalls instances.
>
> The default settings are not correct generated as shown by this error.
>
> colvars: Collective variables initialized, 2 in total.
>
> colvars: ------------------------------------------------------------
> ----------
>
> colvars: Initializing a new "harmonicwalls" instance.
>
> colvars: # name = "chi1lw"
>
> colvars: # colvars = { chi1 }
>
> colvars: # outputEnergy = off [default]
>
> colvars: # forceConstant = 20
>
> colvars: # targetForceConstant = 0 [default]
>
> colvars: # lowerWalls = { -180 }
>
> colvars: # upperWalls = { 0 } [default]
>
> colvars: Upper walls were not provided.
>
> colvars: Error: at least one variable is periodic, both walls must be
> provided .
>
> FATAL ERROR: Error in the collective variables module: exiting.
>
>
>
> This is not described in the namd docs that I can find
>
> Brian
>
>
>
>
>
> *From:* Giacomo Fiorin [mailto:giacomo.fiorin_at_gmail.com]
> *Sent:* Thursday, February 23, 2017 12:44 PM
> *To:* Bennion, Brian <bennion1_at_llnl.gov>
> *Cc:* namd-l_at_ks.uiuc.edu; Ajasja Ljubetič <ajasja.ljubetic_at_gmail.com>
>
> *Subject:* Re: namd-l: colvars: Error: Feature unavailable:compute total
> collective force only from one group center
>
>
>
> Hi Brian, you have defined the keywords lower/upper walls, which will
> confine the variables by half-harmonic potentials, but not the boundaries
> of the PMF grid (lowerBoundary / upperBoundary).
>
> Giacomo
>
>
>
> On Thu, Feb 23, 2017 at 3:35 PM, Bennion, Brian <bennion1_at_llnl.gov> wrote:
>
> Hello,
>
>
>
> I have just recompiled namd2.12 with the cvs nightly from Feb 23.
>
> There have been a number of deprecations that I fixed in my colvar input
> file. However, one error remains that I have not figure out yet.
>
> Below is my current colvar file. I just wat to map the potential energy
> surface for two dihedrals of an amino acid in a protein, Chi1 and Chi2.
>
>
>
> The error is:
>
> colvars: Collective variables initialized, 2 in total.
>
> colvars: ------------------------------------------------------------
> ----------
>
> colvars: Initializing a new "metadynamics" instance.
>
> colvars: # name = "metadynamics1" [default]
>
> colvars: # colvars = { chi1, chi2 }
>
> colvars: # outputEnergy = off [default]
>
> colvars: # hillWeight = 0.5
>
> colvars: # newHillFrequency = 3000
>
> colvars: # hillWidth = 2
>
> colvars: Half-widths of the Gaussian hills (sigma's):
>
> colvars: chi1: 10
>
> colvars: chi2: 10
>
> colvars: # multipleReplicas = off [default]
>
> colvars: # useGrids = on [default]
>
> colvars: # gridsUpdateFrequency = 3000 [default]
>
> colvars: # rebinGrids = off [default]
>
> colvars: Feature unavailable: "lower boundary" in colvar chi1
>
> colvars: ...required by "grid" in colvar chi1
>
> colvars: Error: Failed dependency in colvar chi1.
>
> FATAL ERROR: Error in the collective variables module: exiting.
>
> colvarsTrajFrequency 500
>
> colvarsRestartFrequency 500
>
>
>
>
>
> colvar {
>
> name chi1
>
> width 10.0
>
> lowerWall -180
>
> upperWall 180
>
> lowerWallConstant 0.2
>
> upperWallConstant 0.2
>
> outputAppliedForce on
>
> dihedral {
>
> oneSiteSystemForce
>
> group1 {
>
> atomnumbers 3487
>
> }
>
> group2 {
>
> atomnumbers 3489
>
> }
>
> group3 {
>
> atomnumbers 3491
>
> }
>
> group4 {
>
> atomnumbers 3494
>
> }
>
> }
>
> }
>
>
>
> colvar {
>
> name chi2
>
> width 10.0
>
> lowerWall -180
>
> upperWall 180
>
> lowerWallConstant 0.2
>
> upperWallConstant 0.2
>
> outputAppliedForce on
>
> dihedral {
>
> oneSiteTotalForce
>
> group1 {
>
> atomnumbers 3489
>
> }
>
> group2 {
>
> atomnumbers 3491
>
> }
>
> group3 {
>
> atomnumbers 3494
>
> }
>
> group4 {
>
> atomnumbers 3495
>
> }
>
> }
>
> }
>
>
>
> metadynamics {
>
> colvars chi1 chi2
>
> hillWeight 0.5
>
> hillWidth 2
>
> newHillFrequency 3000
>
> }
>
>
>
> *From:* owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] *On
> Behalf Of *Bennion, Brian
> *Sent:* Thursday, February 23, 2017 8:39 AM
> *To:* Giacomo Fiorin <giacomo.fiorin_at_gmail.com>; NAMD list <
> namd-l_at_ks.uiuc.edu>; Ajasja Ljubetič <ajasja.ljubetic_at_gmail.com>
>
>
> *Subject:* Re: namd-l: colvars: Error: Feature unavailable:compute total
> collective force only from one group center
>
>
>
> Thank you both for the info.
>
> I am using the NAMD_CVS-2016-11-11_Source for my build. I will compile a
> later version.
>
>
>
> Brian
>
>
> ------------------------------
>
> *From:* Giacomo Fiorin <giacomo.fiorin_at_gmail.com>
> *Sent:* Thursday, February 23, 2017 1:10:20 AM
> *To:* NAMD list; Ajasja Ljubetič
> *Cc:* Bennion, Brian
> *Subject:* Re: namd-l: colvars: Error: Feature unavailable:compute total
> collective force only from one group center
>
>
>
> Ajasja pointed correctly to the fix for that bug.
>
> More to that, I believe that the fix is already included in NAMD 2.12: can
> you try it?
>
> Giacomo
>
>
>
> On Thu, Feb 23, 2017 at 2:36 AM, Ajasja Ljubetič <
> ajasja.ljubetic_at_gmail.com> wrote:
>
> Hi!
>
> I'm not 100% sure this is related to your problem, but this was on of the
> larger changes to the colvars module:
> https://colvars.github.io/totalforce.html
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__colvars.github.io_totalforce.html&d=DwMFaQ&c=8hUWFZcy2Z-Za5rBPlktOQ&r=6cQ04kjNHUuxT8SDe1wb0TtPDsnWRp4T4iwoYRLFpsY&m=Nc4rBAmwF0p9LnBBNhMpRiAXvXqyIA8RNz8Kk2kCn90&s=yCbl4Q5dCuBZ-92PFxhp0iPz2bkq8Tf61uqyleuQKmw&e=>
>
> Also, there is one hit with this error here:
> https://github.com/colvars/colvars/issues/89
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_colvars_colvars_issues_89&d=DwMFaQ&c=8hUWFZcy2Z-Za5rBPlktOQ&r=6cQ04kjNHUuxT8SDe1wb0TtPDsnWRp4T4iwoYRLFpsY&m=Nc4rBAmwF0p9LnBBNhMpRiAXvXqyIA8RNz8Kk2kCn90&s=X4Iku8BTWbq1vhFT_wKafCW2luDahtDtzy8ZY1rChpk&e=>
>
>
>
> Best,
>
> Ajasja
>
>
>
> On 23 February 2017 at 01:17, Bennion, Brian <bennion1_at_llnl.gov> wrote:
>
> Hello,
>
> Getting back to some work I did last year with metadynamics. Hardware has
> been updated since then as well as the version of namd.
>
> I have scanned the email list for the error described below but have not
> found anything yet. I know that there has been a fair amount of work in
> the colvars modules in between namd.10 and namd2.12. Perhaps I just need
> to alter a keyword. The colvar output is below:
>
>
>
> colvars: ------------------------------------------------------------
> ----------
>
> colvars: Initializing the collective variables module, version 2016-10-21--001a1148494a9ea1cb05493ea2ce--

This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:21:06 CST