[gmx-developers] Large VCM(group rest): Temp-cm: inf

[Carsten Kutzner]

Am 18.09.2008 um 17:43 schrieb Berk Hess:

> Hi,
>
> I have added a check on group_ndf.
>
> But how can you have a vcm group with non-zero mass, but with zero  
> degrees of freedom?
>
Hi,

In the mdp file just a single comm-grp was defined, but in the  
debugger I saw that
the number of groups is 2, with the second group having no name.  
Maybe thte number
of groups should be 1 in that case anyway?

Carsten

> Berk
>
> Carsten Kutzner wrote:
>> Hi,
>>
>> we appear to have another small problem with the CVS version.
>> For a protein in water system, where comm-grps=Protein is
>> chosen we continuously get the following line outputted to log:
>>
>> Large VCM(group rest):      0.00005,     -0.00002,     -0.00004,  
>> Temp-cm:          inf
>> Large VCM(group rest):      0.00006,     -0.00003,     -0.00004,  
>> Temp-cm:          inf
>> Large VCM(group rest):      0.00006,     -0.00003,     -0.00003,  
>> Temp-cm:          inf
>> Large VCM(group rest):      0.00006,     -0.00004,     -0.00002,  
>> Temp-cm:          inf
>> Large VCM(group rest):      0.00006,     -0.00005,     -0.00001,  
>> Temp-cm:          inf
>> ...
>>
>> This happens also for the non-MPI mdrun. The problem is that in  
>> check_cm_grp (vcm.c)
>> the number of degrees of freedom for the 2nd group are zero. Thus
>>
>> Temp_cm = 2*ekcm/vcm->group_ndf[g];
>>
>> becomes infinity, triggering the message. This is just an output  
>> problem, since
>> nothing needs to be done to the COM of the second group anyway.  
>> However, the md.log files
>> become HUGE when this is outputted at every step.
>>
>> Carsten
>>
>>
>>
>>
>
> _______________________________________________
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the www  
> interface or send it to gmx-developers-request at gromacs.org.