[gmx-users] RE: g_anaeig output

[Berk Hess]

>From: akdemir at few.vu.nl
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: gmx-users at gromacs.org
>Subject: [gmx-users] RE: g_anaeig output
>Date: Tue, 15 Jun 2004 15:00:54 +0200
>
>Dear Berk (and others)
>
>This is what I mean.
>
>I performed g_anaeig on the eigenvectors and eigenvalues that I obtained 
>from
>g_covar. I have chosen the option -comp -rmsf and -proj. -last and -first 
>are
>default so it gives the output for the first until the eight eigenvector. 
>The
>output of -proj will give the eigenvalue of an eigenvector as a function of
>the time (if I understood correctly). The problem is that eigenvector 1 
>only
>spans half of the total simulation time, while the other vectors span the
>whole simulation time.

-proj gives the projection on the eigenvector (not the eigenvalue) as a 
function
of time. The average of the square of this projection is the eigenvalue.

>
>I also see a related problem for the output of -comp and -rmsf. Here the 
>value
>of the eigenvectors as a function of the atomnumber is given as output.
>However, I only see half of my atoms in eigenvector 1, but all atoms in the
>other eigenvectors.

It would seem to me that there is somehow some inconsistency in the input
for g_anaeig. The only thing I can think of now is that the eigenvec.trr 
file
is interpreted incorrectly. But I don't know how this could have happened.

Which Gromacs version do you use?
You did not change the eigenvec.trr file in any way?
Could you mail the exact command line options you used with g_covar and 
g_anaeig?
Which groups with how many elements did you select in g_covar and g_anaeig?

Berk.

_________________________________________________________________
Stop worrying about overloading your inbox - get MSN Hotmail Extra Storage! 
http://join.msn.click-url.com/go/onm00200362ave/direct/01/