RE: Questions about lattice information in NAMD2.7b2

From: Bennion, Brian (Bennion1_at_llnl.gov)
Date: Fri Feb 12 2010 - 12:42:45 CST

Hell Jim,

Doh... didn't see that at even though I had that file wide open on the monitor.

Thanks that did the trick. Added "friend class GlobalMasterBabak" to NamdState.h.
Now I have realtime lattice information in my GlobalMasterBabak class.
Nice...

Thank you for the help.

Brian
-----Original Message-----
From: Jim Phillips [mailto:jim_at_ks.uiuc.edu]
Sent: Friday, February 12, 2010 10:24 AM
To: Bennion, Brian
Subject: RE: Questions about lattice information in NAMD2.7b2

"friend class Controller" in NamdState.h

-Jim

On Fri, 12 Feb 2010, Bennion, Brian wrote:

> Hello Jim,
> Thank you for the very quick comment. This is what I have done with your suggestion
>
> 1) added Node.h and NamdState.h to includes in my GlobalMasterBabak.C
> 2) added line Lattice *lattice = Node::Object()->state->lattice;
> 3) compiler states that "member "NandState::lattice" (declared at line 33 in src/NamdState.h") is inaccessible.
> That's because it is sitting in the private section in that declaration.
> How in the world does the other code in Controller.C see this information...
>
> Brian
>
> -----Original Message-----
> From: Jim Phillips [mailto:jim_at_ks.uiuc.edu]
> Sent: Thursday, February 11, 2010 5:13 PM
> To: Bennion, Brian
> Subject: Re: Questions about lattice information in NAMD2.7b2
>
> Hi,
>
> Try Node::Object()->state->lattice. Let me and namd-l know if it works.
>
> -Jim
>
>
> On Thu, 11 Feb 2010, Bennion, Brian wrote:
>
>> Hello Jim,
>>
>> I hope you are well. Recently I posted a query to the namd-l list and
>> it was probably to deep for the general readership.
>>
>> Could you tell me how I could access current lattice information in a
>> class derived from the GlobalMaster class?
>>
>> I have a tool that I added to NAMD in similar fashion to the colvar and
>> iAPBS routines in the GlobalMaster class. However, I have to account
>> for people using the langevin piston which changes the box dimensions at
>> very small increments each step which also scales the atom positions by
>> some small factor. The final state in our tool needs to also be scaled
>> by the same factor
>>
>> Looked at Output.C and Controller.C to see how they access the lattice
>> information, but there are at least a half a dozen times where the
>> lattice object is passed and referenced and dereferenced.
>>
>> So far I am only currently able to see the starting dimensions as found
>> in SimParameters and not updated information.
>>
>> Thank you for your time and any help.
>>
>> Brian Bennion
>> Biosciences and Biotechnology Division
>> Lawrence Livermore National Laboratory
>> 7000 East Ave Livermore California 94550
>> phone:925-422-5722
>> fax:925-424-4224
>>
>>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:22:43 CST