[gmx-developers] correct usage of do_fit

[Erik Lindahl]

Hi,

On Mar 4, 2007, at 7:53 PM, Martin Höfling wrote:

>
> ... and gromacs is well tested too :-) . My fault was the mass  
> vector. I
> provided only the masses of the fit group althoug it needs the full  
> array (
> probably only using the values in fit index ).

Correct - sorry if that wasn't documented. The reason for using the  
full array (i.e. atomic indices) is that you sometimes fit one group  
and then calculate RMSD for 10 different index groups. In that case  
it seems stupid to have to create 10 different mass arrays, just  
because the index groups are different.

Cheers,

Erik