[gmx-developers] correct usage of do_fit

[Erik Lindahl]

Hi Martin,

This means it was not possible to diagonalize the matrix - the  
iteration limit is set very high, so there is no realistic scenario I  
can think of where it needs to be higher.

I would try writing out the (two) coordinate sets you are trying to  
fit, look at them in PyMOL (just to make sure there are no horrible  
artifacts), and then try to fit them e.g. with vanilla trjconv or g_rms.

If that does not work either, then it could be a bug in the routine,  
but this is a pretty well-tested part of Gromacs, so hopefully that's  
unlikely :-)

Cheers,

Erik

On Mar 3, 2007, at 3:58 PM, Martin Höfling wrote:

> Hi there,
>
> i am getting a Fatal error:
> Error: Too many iterations in routine JACOBI
>
> I am trying to output the rotation matrices when rotating a part of  
> the
> trajectory using gmx_trjconv as a template for my program.
>
> First step was rewriting do_fitWO(natoms,masses,xorig,x,out_file);  
> as a copy
> of do_fit with additional output.
>
> After the usual stuff like setting up and reading first frame, i do  
> the
> following:
>
> /*Storing masses*/
> fprintf(stderr,"\nStoring masses of %d atoms\n",isize[0]);
> real *masses;
> snew(masses,isize[0]);
> for (i=0; i<isize[0];i++) {
>         	j=index[0][i];
>         	masses[i]=top.atoms.atom[j].m;
> }
>
> /*prepare and center copy of first frame*/
> snew(xorig,natoms);
> memcpy(xorig,x,sizeof(rvec)*natoms);
> reset_x(isize[0],index[0],natoms,NULL,xorig,masses);	
>
> /*Main loop over frames*/
> while(1) {
> 		/*remove com position*/
> 		reset_x(isize[0],index[0],natoms,NULL,x,masses);
>
> 		/*do fit with writeout*/	
> 		do_fitWO(natoms,masses,xorig,x,out_file);
>
> 		/*try to read next frame*/	
>                 if ( read_next_x ( status,&t,natoms,x,box ) ==FALSE )
> 	            break;
> }
>
> What could I have done wrong?
>
> Cheers
> 	Martin
> _______________________________________________
> 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.