Re: Multi canonical ensemble method

From: Giacomo Fiorin (giacomo.fiorin_at_gmail.com)
Date: Wed Nov 06 2019 - 13:32:45 CST

Hello Takeru, the dependency error is due to trying to measure a total
force for a variable type that doesn't have it implemented. Below is the
updated link, which unlike the one you copied earlier, should remain the
same in future versions (and thus be useful in searches for this message):
https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:cvc_sys_forces

You can still use ABF on an extended-Lagrangian variable, but you probably
need to get rid of outputTotalForce. Does that make solve the problem?

The second message about sockets is bogus, sorry. For a while (but not any
longer) Colvars has been sharing the error code flag with other functions
in NAMD, but it wasn't always set consistently and the numeric code
sometimes referred to something completely unrelated, like that. In
general, the Colvars module will always print all errors originating there
first, and the last "FATAL ERROR" message from the main NAMD function
simply refers to that, but does not add any further information.

Giacomo

On Wed, Nov 6, 2019 at 4:58 AM Takeru KAMEDA <kamedapon_at_hiroshima-u.ac.jp>
wrote:

> Dear Giacomo.
>
> Thank you for your kind reply.
>
> And, I will take care to keep the list in CC.
>
> I changed parentheses to curly braces as :
>
> axis {1, 0, 0} → axis (1, 0, 0)
>
>
> Then, following error message were output :
>
> Excerpt below
>
>
> """"""""
> colvars:
> ----------------------------------------------------------------------
> colvars: Collective variables initialized, 1 in total.
> colvars:
> ----------------------------------------------------------------------
> colvars: Initializing a new "abf" instance.
> colvars: # name = "abf1" [default]
> colvars: # colvars = { theta }
> colvars: # outputEnergy = off [default]
> colvars: # timeStepFactor = 1 [default]
> colvars: # applyBias = on [default]
> colvars: # updateBias = on [default]
> colvars: # hideJacobian = on
> colvars: Jacobian (geometric) forces will be handled internally.
> colvars: # fullSamples = 200
> colvars: # inputPrefix = [default]
> colvars: # outputFreq = 50000 [default]
> colvars: # historyFreq = 0 [default]
> colvars: # shared = off [default]
> colvars: "total force" in colvar theta requires one of the following
> features, none of which can be enabled:
> colvars: -----------------------------------------
> colvars: 1. extended Lagrangian
> colvars: User-controlled feature "extended Lagrangian" may not be
> enabled as a dependency in colvar theta.
> colvars: 2. total force calculation
> colvars: Dynamic feature unavailable: "Jacobian derivative" in
> cvc tilt0001.
> colvars: ...required by "Jacobian derivative" in colvar theta
> colvars: ...required by "total force calculation" in colvar theta
> colvars: -----------------------------------------
> colvars: ...required by "obtain total force" in bias abf1
> colvars: Error: Failed dependency in bias abf1.
> FATAL ERROR: Error in the collective variables module: Socket operation on
> non-socket
> """"""""
>
> Are there 2 problems ?
>
>
> In previous exchange for others, I obtained following manual page
>
>
> http://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#x1-530005.3.3
>
> But, I cloudn't resolve above problem
>
> Thanks.
>
> Best wishes,
> Takeru Kameda
> PhD Student, Hiroshima University, Japan
>
> ________________________________________
> 差出人: Giacomo Fiorin <giacomo.fiorin_at_gmail.com>
> 送信日時: 2019年11月5日 22:18
> 宛先: Takeru KAMEDA
> CC: NAMD list
> 件名: Re: namd-l: Multi canonical ensemble method
>
> Hello Takeru, please keep the list in CC.
>
> Your error is due to using curly braces instead of parentheses.
>
> Giacomo
>
> On Mon, Nov 4, 2019 at 9:39 PM Takeru KAMEDA <kamedapon_at_hiroshima-u.ac.jp
> <mailto:kamedapon_at_hiroshima-u.ac.jp>> wrote:
> Dear Giacomo
>
> Thank you for your kind reply.
>
> I tried ABF as you said using attached files.
>
> However, I obtained attached log file (run.log).
>
> Are there any missing configurauin ?
>
> Thanks
>
> Best wishes,
> Takeru Kameda
> PhD Student, Hiroshima University, Japan
>
> ________________________________________
> 差出人: Giacomo Fiorin <giacomo.fiorin_at_gmail.com<mailto:
> giacomo.fiorin_at_gmail.com>>
> 送信日時: 2019年10月30日 23:22
> 宛先: Takeru KAMEDA
> CC: Jérôme Hénin; namd-l
> 件名: Re: namd-l: Multi canonical ensemble method
>
> Hello Takeru, it looks like you already found the documentation, try to
> double-check that your configuration is consistent. See here an example:
> https://raw.githubusercontent.com/Colvars/colvars/master/tests/spinangle.in
> <
> https://jpn01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fraw.githubusercontent.com%2FColvars%2Fcolvars%2Fmaster%2Ftests%2Fspinangle.in&data=02%7C01%7Ckamedapon%40hiroshima-u.ac.jp%7C99bff67013fc40e50bd008d761f2e924%7Cc40454ddb2634926868d8e12640d3750%7C1%7C0%7C637085568234920878&sdata=soTnsL1ev1023QcQwGrqktuaU1TLk92cFx00B3PPUwA%3D&reserved=0
> ><
> https://jpn01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fraw.githubusercontent.com%2FColvars%2Fcolvars%2Fmaster%2Ftests%2Fspinangle.in&data=02%7C01%7Ckamedapon%40hiroshima-u.ac.jp%7Cdc8ab578dcfd4144be5308d75d44c110%7Cc40454ddb2634926868d8e12640d3750%7C1%7C0%7C637080422187028882&sdata=uDrPerxN1ceCjrQhZ4yfQ9Qpp3xln2AQwhnE4Cs%2B5ak%3D&reserved=0
> <
> https://jpn01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fraw.githubusercontent.com%2FColvars%2Fcolvars%2Fmaster%2Ftests%2Fspinangle.in&data=02%7C01%7Ckamedapon%40hiroshima-u.ac.jp%7C99bff67013fc40e50bd008d761f2e924%7Cc40454ddb2634926868d8e12640d3750%7C1%7C0%7C637085568234920878&sdata=soTnsL1ev1023QcQwGrqktuaU1TLk92cFx00B3PPUwA%3D&reserved=0
> >>
>
> If it is consistent with the doc, do you mind pasting it in the email or
> attaching it?
>
> Giacomo
>
> On Wed, Oct 30, 2019 at 1:28 AM Takeru KAMEDA <kamedapon_at_hiroshima-u.ac.jp
> <mailto:kamedapon_at_hiroshima-u.ac.jp><mailto:kamedapon_at_hiroshima-u.ac.jp
> <mailto:kamedapon_at_hiroshima-u.ac.jp>>> wrote:
> Hi Jerome.
>
> Thank you for your advice.
> And, sorry for the late reply.
>
>
> I changed the method orientation to spinAngle.
> https://www.ks.uiuc.edu/Research/namd/2.13/ug/node55.html#24897<
> https://jpn01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.ks.uiuc.edu%2FResearch%2Fnamd%2F2.13%2Fug%2Fnode55.html%2324897&data=02%7C01%7Ckamedapon%40hiroshima-u.ac.jp%7C99bff67013fc40e50bd008d761f2e924%7Cc40454ddb2634926868d8e12640d3750%7C1%7C0%7C637085568234930867&sdata=D88d4MWTzKi%2Fw2Atyjk5rNhWkB5%2BqEQxH7kx2M3a2pc%3D&reserved=0
> ><
> https://jpn01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.ks.uiuc.edu%2FResearch%2Fnamd%2F2.13%2Fug%2Fnode55.html%2324897&data=02%7C01%7Ckamedapon%40hiroshima-u.ac.jp%7Cdc8ab578dcfd4144be5308d75d44c110%7Cc40454ddb2634926868d8e12640d3750%7C1%7C0%7C637080422187028882&sdata=XY27Nbbf20n3dSyNa%2BJOc7key7RYeJHD3602iGzuaMU%3D&reserved=0
> <
> https://jpn01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.ks.uiuc.edu%2FResearch%2Fnamd%2F2.13%2Fug%2Fnode55.html%2324897&data=02%7C01%7Ckamedapon%40hiroshima-u.ac.jp%7C99bff67013fc40e50bd008d761f2e924%7Cc40454ddb2634926868d8e12640d3750%7C1%7C0%7C637085568234930867&sdata=D88d4MWTzKi%2Fw2Atyjk5rNhWkB5%2BqEQxH7kx2M3a2pc%3D&reserved=0
> >>
>
> Now I'm trying such method.
> However, I can not set "axis" part in "List of keywords" of spinAngle.
>
> In this case, I think I must set the axis of rotation and center of
> roration.
>
> Do you know how to set such axis and center.
>
>
> Best wishes,
> Takeru Kameda
> PhD Student, Hiroshima University, Japan
>
> ________________________________________
> 差出人: Jérôme Hénin <jerome.henin_at_ibpc.fr<mailto:jerome.henin_at_ibpc.fr
> ><mailto:jerome.henin_at_ibpc.fr<mailto:jerome.henin_at_ibpc.fr>>>
> 送信日時: 2019年9月30日 20:49
> 宛先: Takeru KAMEDA
> CC: namd-l
> 件名: Re: namd-l: Multi canonical ensemble method
>
> Dear Takeru,
>
> this "failed dependency" message indicates that ABF only supports scalar
> variables. A multidimensional ABF calculation can be done with several
> scalar variables, not a single vector one.
>
> In this case, the orientation variable is 4-dimensional, which is probably
> too much for ABF anyway. I recommend reducing your sampling space to 1, 2
> or 3 scalar quantities.
>
> Best wishes,
> Jerome
>
> On Mon, 30 Sep 2019 at 04:23, Takeru KAMEDA <kamedapon_at_hiroshima-u.ac.jp
> <mailto:kamedapon_at_hiroshima-u.ac.jp><mailto:kamedapon_at_hiroshima-u.ac.jp
> <mailto:kamedapon_at_hiroshima-u.ac.jp>><mailto:kamedapon_at_hiroshima-u.ac.jp
> <mailto:kamedapon_at_hiroshima-u.ac.jp><mailto:kamedapon_at_hiroshima-u.ac.jp
> <mailto:kamedapon_at_hiroshima-u.ac.jp>>>> wrote:
> Dear all.
>
> Hi. I am trying multi-canonical ensemble method using NAMD2.13.
>
> I wrote colvar file as follows:
>
> colvarsTrajFrequency 5000
> colvarsRestartFrequency 50000
>
> colvar {
> name theta
>
> outputValue on
> #outputVelocity on
> #outputEnergy on
> #outputSystemForce on
> #outputAppliedForce on
>
> orientation {
> atoms { atomnumbers {
> 13754 13755 13756 13757 13758 13761 13783 13785 19290
> 19291 19292 19293 19294 19297 19319 19321
> }}
> refPositionsFile ./colvar.pdb
> refPositionsCol O
> }
> }
>
> abf {
> colvars theta
> fullSamples 200
> hideJacobian
> }
>
>
> Then, we execute simulation, following error was output:
>
> colvars:
> ----------------------------------------------------------------------
> colvars: Initializing the collective variables module, version 2018-10-16--000000000000a803220596b2aa60--

This archive was generated by hypermail 2.1.6 : Sat Dec 07 2019 - 23:20:53 CST