[Ifeffit] choosing dk values

[Matt Newville]

Hi Mike,

On Mon, 1 Nov 2004, Michael A Groves wrote:

> I'm doing some fitting with Artemis, and I noticed that, for
> different values of dk, I get better or worse fits.  The only
> information I've gotten so far about dk was in the IFEFFIT
> documentation, and it basically said, "Don't mess with it."  
> Can anyone give me more information on choosing a good dk value?  
> Or, should I just stick with the default?

Hmm, I'm not sure where that is said in documents.  Dk is the size
of the "window sill", and gives the extent of k over which the FT
windows rises.  Thus it can change the k-range of the data, and
definitely has an effect on the EXAFS FT.  It can have an impact
on the fit results, but I'd hope any impact is not too big!

For many window (notably the Hanning window) types, Dk is the
k-extent over which the window function goes from 0 to 1.  For
some windows (notably the Kaiser-Bessel window), the window
function never goes to zero, so the meaning of Dk is somewhat
harder to visualize.  Some plots of typical FT window functions,
and a little more discussion is at
 http://cars9.uchicago.edu/ifeffit/refman/node142.html

I tend to use either Hanning windows with dk between 1 and 2, or I
use the Kaiser-Bessel window with dk between 3 and 5.  I'd go as
far as saying I recommend these values, but I'm sure others have
different and perfectly valid opinions.  Anyway, dk is definitely
something you can play with!

Hope that helps,

--Matt