[Ifeffit] Re: Artemis question

Bruce Ravel ravel at phys.washington.edu
Thu Jan 22 12:17:50 CST 2004

On Thursday 22 January 2004 11:57 am, Matt Newville wrote:
> I agree with most of your answer to Dan about setting constraints
> and restraints, but I think I'd fail the 'quiz' you gave:
> > A word of caution.  Restraints can be dangerous.  Here is a quiz:  Why
> > is this definition of the restraint a bad idea?
> >
> >       set      scale = -100
> >       restrain amp   = abs(amp_guess - amp_target) * scale
> where you had
> >     guess    amp_guess  = 0.85
> >     set      amp_target = 0.85
> >     set      scale      = 100
> >     restrain amp        = abs(amp_guess - amp_target) * scale
> as a recommended way to set just such a constraint. I believe
> there's no difference.  Am I missing something?

No... probably I am missing something.

It seems to me that if you use the absolute value to force the
difference to be a positive number, then set the scale to a negative
number, that would *reward* a large difference rather than *penalize*

But perhaps I am confused in how the restraints are added to
chi-square.  I am under the impression that restraints are added as
written to chi-square.  This seems consistent with the language in
section 7.9 of the reference manual.  However, if the restraint is
squared before adding to chi-square then what I said about the
negative value of the scaling term is clearly wrong.

I should point out that I have not actually looked at the source


 Bruce Ravel  ----------------------------------- ravel at phys.washington.edu
 Code 6134, Building 3, Room 222
 Naval Research Laboratory                          phone: (1) 202 767 5947
 Washington DC 20375, USA                             fax: (1) 202 767 1697

 NRL Synchrotron Radiation Consortium (NRL-SRC)
 Beamlines X11a, X11b, X23b
 National Synchrotron Light Source
 Brookhaven National Laboratory, Upton, NY 11973

 My homepage:    http://feff.phys.washington.edu/~ravel 
 EXAFS software: http://feff.phys.washington.edu/~ravel/software/exafs/

More information about the Ifeffit mailing list