AMBER Archive (2009)

Subject: Re: [AMBER] projections of trajectories onto the first 2 principal eigenvectors

From: Hannes Loeffler (
Date: Fri Feb 13 2009 - 03:33:19 CST

On Fri, 2009-02-13 at 08:32 +0000, Hannes Loeffler wrote:
> I do not understand your last sentence. A projection is a dot product
> between two vectors naturally yielding a scalar. I do not see how you
> could possibly obtain vectors in this way.

On second thought I may understand what you are after. Do you view the
two projections themselves as vectors? You won't get a matrix in this
way but you can plot corresponding pairs in a 2D plot and thereby you
will get a "matrix". (When you connect all individual points according
to its time relationship you will obviously get the trajectory projected
down to two dimensions only.) That is what some people do in their

> On Thu, 2009-02-12 at 12:22 -0600, Carra, Claudio (JSC-SK)[USRA] wrote:
> > Dear All,
> > is there a way in amber to calculate the projections
> > of trajectories onto the first 2 principal eigenvectors?
> > I'm interested in looking at the distribution densities.
> > I would expect a matrix but if I use "projection"
> > I get 2 vectors.
> _______________________________________________
> AMBER mailing list

AMBER mailing list