[CP2K:1441] No TARGET field in FIXED_ATOMS section

[Laino Teodoro]

section CONSTRAINT%FIXED_ATOMS can NOT have a TARGET since there you  
may specify a list of atoms..
and a single target would be meaningless unless you don't want all of  
them collapse in one point/plane/line (according
the amount of components you want to keep fixed).
A more complicated TARGET (like multiple TARGET or list of numbers is  
out of question.. far too complicated and error
prone..)
So.. there cannot be a TARGET keyword in FIXED.

Moreover normally you start from a geometry where you want to keep  
frozen/restraint some components of few atoms..
if you want to do something more exotic.. of course you can..
there's a section:

http://cp2k.berlios.de/manual/CP2K_INPUT/MOTION/CONSTRAINT/ 
FIX_ATOM_RESTART.html

where normally are stored the coordinates at the zero step  
(considered as the TARGET).
you can use that section to do what you want to do..

Teo

On 14 Oct 2008, at 18:33, ilya wrote:

>
> Hi!!
>
> I wonder why there is no TARGET field in CP2K_INPUT / MOTION /
> CONSTRAINT / FIXED_ATOMS section, though CP2K_INPUT / MOTION /
> CONSTRAINT / COLLECTIVE does have such field. As I understand it uses
> initial positions from the input. But how to constrain atoms to the
> slightly different positions then? By the way comment to the
> CP2K_INPUT / MOTION / CONSTRAINT / FIXED_ATOMS / RESTRAINT / K   field
> makes a reference to the TARGET: "The functional form for the
> restraint is: K*(X-TARGET)^2".
>
> Sincerely,
> Ilya.
> >