[Ifeffit] the use of constraints

[Matt Newville]

Hi Kellie,

It's a little hard to tell for sure why your "percap" is going
slightly greater than 1 without knowing more details of the analysis.
 Questions like the following come to mind:
       Did you include an S02 term?
       Is "focc" fixed?  If so, could the value be slightly off?
       Is sigma2 set or being adjusted -- could this be offf?

I wouldn't expect direct answers to these, just pointing them out as
things to re-check.   There are a lot of terms that effect the XAFS
amplitude, and it can be tricky when you're trying to de-couple them.

Cheers,

--Matt




On Fri, May 28, 2010 at 11:01 AM, Kellie Aldi <kellie.morgan at gmail.com> wrote:
> Hi, I have a question on the use of constraints and restraints in EXAFS data
> fitting.
>
> I am using SixPack to build a model for the EXAFS of some synthetic layered
> transition metal (TM) oxides with relatively large amounts of layer
> vacancies. I am refining the occupancy of the layer TM site, from which I
> can determine the amount of layer vacancies. I want to include a parameter
> in this model that refines the number of these vacancies that are capped by
> interlayer TM (something that can happen in these structures), but I am
> having difficulty in deciding a responsible way to define the parameters of
> this variable. I am modifying the intensity of the interlayer TM scattering
> paths with a parameter
>
> fcap = (1-focc)*percap
>
> where focc is the refined occupancy of the layer TM site, and percap is the
> percentage of vacancies, "1-focc", that are capped by interlayer TM.
>
> However, just letting percap float, the value goes greater that 1... which,
> as a percentage, does not make much sense physically in my model. So I want
> to keep the value of this between 0 and 1. I have tried what was suggested
> on the IFEFFIT frequently asked questions page regarding constraints and
> restraints. I don't feel comfortable simply using a constraint and forcing
> the value to be between 0 and 1 using the suggested min/max idiom.
>
> I like better the idea of penalizing a fit when percap is outside of a
> defined range and thus "encouraging" the fit to settle percap somewhere
> inside this range. But in trying to use the restraint idea, including the
> fix suggested by Matt Newville for only penalizing the fit outside of the
> min/max range, percap still goes slightly above 1 (about 1.02) and in my
> results I get a value for the "penalty" that I am unsure how to interpret.
> Can this value be used to understand why the restraint still allows the
> value of percap to be outside of the defined "ideal" range?
>
> I am trying to understand how the restraint that Matt Newville defined in
> the link in the restraints/constraints FAQ works, but I am relatively new to
> EXAFS modeling and thus shaky on my skills at defining these parameters and
> knowing how certain math functions can be used. Are there any other help
> sections or guides that might address the use of such restraints? Or any
> suggestions on how to re-define this restraint?
>
> I would greatly appreciate any suggestions/input!
>
> Thanks,
>
> Kellie Aldi
>
> _______________________________________________
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
>
>