[gmx-developers] Regd. potential on grid caculation in PME

[Erik Lindahl]

Hi,

Have a look in src/mdlib/pme.c, routine gather_f_bsplines(). The 
potential at the points is available through the th{x/y/z} arrays, and 
the derivative in dth{x/y/z}.

It looks somewhat complicated due to the 4th order interpolation. The 
best source of documentation on it is probably the Darden smooth PME 
paper (Essman, 1995).

Cheers,

Erik


On Nov 9, 2004, at 5:26 PM, Anjan Raghunathan wrote:

> Hi Erik,
> Thank you for the observation. Actually I would not mind having the 
> reciprocal space potential, as could calculate direct space rather a 
> bit tediously. And am only interested in Electrostatics so if I am not 
> wrong I dont need the 1,4 interactions. Any suggestions on how to get 
> the reciprocal space contribution that are calculated on the grids in 
> pme.c?
>
> thanks,
> Anjan
>
> Erik Lindahl wrote:
>
>> Hi Anjan,
>>
>> I don't think it will work to use the PME grid - remember that it 
>> only contains the reciprocal space contribution to the potential, not 
>> the direct-space part or any 1,4 contributions.
>>
>> Cheers,
>>
>> Erik
>>
>>
>> On Nov 3, 2004, at 9:53 PM, Anjan Raghunathan wrote:
>>
>>> Hi all,
>>> Actually I went back to the pme.c file to find where the potentials 
>>> were being calculated. I believe they are calculated on a grid sent 
>>> to fft with arguments nx,ny,nz. Unfortunately I am not well versed 
>>> with fftw , so if some one could point me out to how I could use the 
>>> grids potential calculated there to get the grids on the real space 
>>> grid I would appreciate it.
>>>
>>> thanks,
>>>
>>> Anjan Raghunathan
>>>
>>> Anjan Raghunathan wrote:
>>>
>>>> Hi all,
>>>> I'd like to know if there is a way of getting the
>>>> electrostatic potentials from PME(direct+k space) in a
>>>> plane(grid) that I define in the system every time step. I
>>>> thought of using dummy atoms but dont know if the energy
>>>> exclusion will take care of excluding the coulomb interactions
>>>> also(usually get warnings like cannot exclude coulomb
>>>> interactions.....).
>>>> I would appreciate it if some one could point me to how
>>>> this could be accomplished or what code could be changed to do
>>>> this.
>>>>
>>>> thanks,
>>>> Anjan Raghunathan
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> 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.
>>>>
>>>
>>> _______________________________________________
>>> 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.
>>
>>
>> _______________________________________________
>> 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.
>>
>
> _______________________________________________
> 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.