Re: Rattle error with lateral pressure calculations

From: Brian Radak (brian.radak_at_gmail.com)
Date: Tue Jul 31 2018 - 12:41:31 CDT

If the energies from stdout and post-processing frames do not match to at
least 3 decimals then I would say that you probably can't trust the results
and something is wrong.

My suspicion is that coorfile does not cause the cell basis vectors to be
updated at each frame, even though that information should be present in
the dcd file. You might have to extract this by some other means (did you
write an xst file by chance?) and then manually reset cellBasisVector1/2/3
inside your while statement.

I've never actually used the pressureProfile code before, so I have no idea
how it is supposed to be used in practice. Does the simulation have to be
constant pressure? Constant volume could solve the problem. You might
consider running a short NVT simulation and then repeating your analysis
procedure. If that solves the problem then I'm probably correct, if not,
then it must be something else entirely.

HTH,
BKR

On Tue, Jul 31, 2018 at 1:11 PM, Sadegh Faramarzi Ganjabad <
safaramarziganjabad_at_mix.wvu.edu> wrote:

> Brian,
>
> I get the error right after the energies appear on the output for the
> first time. The ELECT energy is %30 different from the original trajectory.
>
> Sadegh Faramarzi,
> Research Assistant
> West Virginia University, Department of Chemistry
> Email:safaramarziganjabad_at_mix.wvu.edu
>
> On Mon, Jul 30, 2018 at 9:29 AM, Brian Radak <brian.radak_at_gmail.com>
> wrote:
>
>> If you have dynamics output - do the energies (I would check POTENTIAL
>> and/or ELECT) match when just using a simple coorfile script?
>>
>> On Sun, Jul 29, 2018 at 11:18 PM, Sadegh Faramarzi Ganjabad <
>> safaramarziganjabad_at_mix.wvu.edu> wrote:
>>
>>> Brian,
>>>
>>> Thanks for replying. Yes, the DCD is generated at constant pressure. The
>>> trajectory is wrapped. Is there a way to check if lattice vectors are
>>> updated?
>>>
>>> Best,
>>> Sadegh Faramarzi,
>>> Research Assistant
>>> West Virginia University, Department of Chemistry
>>> Email:safaramarziganjabad_at_mix.wvu.edu
>>>
>>> On Sat, Jul 28, 2018 at 1:17 PM, Brian Radak <brian.radak_at_gmail.com>
>>> wrote:
>>>
>>>> This might just be a wrapping error. Is the DCD generated at constant
>>>> pressure? The lattice vectors might not be getting updated via coorfile.
>>>>
>>>> On Tue, Jul 24, 2018, 6:26 PM Sadegh Faramarzi Ganjabad <
>>>> safaramarziganjabad_at_mix.wvu.edu> wrote:
>>>>
>>>>> Hello,
>>>>>
>>>>> I am doing lateral pressure calculations on a dcd trajectory. Here is
>>>>> the section for pressure profile calculation on NAMD configuration;
>>>>>
>>>>>
>>>>> # set pressure profile parameters
>>>>> pressureProfile on # on to record kinetic, bonded and nb
>>>>> contritbutions
>>>>> pressureProfileSlabs 80 # number of slabs along z
>>>>> pressureProfileFreq 10 # output frequency in timesteps
>>>>> pressureProfileEwald on # when on only Ewald contribution is computed
>>>>> pressureProfileEwaldX 75 # number of vectors along x
>>>>> pressureProfileEwaldY 75 # number of vectors along y
>>>>> pressureProfileEwaldZ 96 # number of vectors along z
>>>>>
>>>>> set ts 0
>>>>> firstTimestep $ts
>>>>>
>>>>> # run pressure calculation
>>>>> coorfile open dcd 1-2Stride100WrappedAligned.dcd
>>>>> while { [coorfile read] != -1 } {
>>>>> incr ts 1000
>>>>> firstTimestep $ts
>>>>> run 0
>>>>> }
>>>>> coorfile close
>>>>>
>>>>> The rest of parameters in the configuration file are pretty much the
>>>>> same as my original MD simulations, except I deleted the minimization and
>>>>> run section. I keep getting rattle error after a couple of timesteps. I
>>>>> thought with this setup I just use the already existing dcd file, rather
>>>>> than running a new simulation that might crash. I'm not sure how to get rid
>>>>> of this error. Do I have to modify my trajectory prior to the calculation?
>>>>>
>>>>> Thanks,
>>>>> Sadegh Faramarzi,
>>>>> Research Assistant
>>>>> West Virginia University, Department of Chemistry
>>>>> Email:safaramarziganjabad_at_mix.wvu.edu
>>>>>
>>>>
>>>
>>
>

This archive was generated by hypermail 2.1.6 : Tue Sep 17 2019 - 23:20:07 CDT