[Ifeffit] path averaging in Artemis

[Ravel, Bruce]

Hi Simon,

That's not a "problem".  Quite the contrary.  It is by design and is something that I consider to be one of the compelling new features in Artemis.

In the new versions of Atoms and Artemis, I introduced a concept called "fuzzy degeneracy" whose purpose is explicitly to combine together paths that differ very little in distance and/or scattering angle.  I explain it in some detail here:

http://bruceravel.github.com/demeter/dpg/feff/pathfinder.html#Fuzzydegeneracy

This feature is entirely user-configurable.  That is, you can set the width of the bin used to group together fuzzily degenerate paths.  In Artemis, this can be set by selecting "Edit Preferences" from the File menu on the main window.  In the preferences window, select "pathfinder" then "fuzz".  The default value is 0.003 Angstroms.  In your case, you want to set this to be 0.01 or smaller if, in fact, you want to retain the split degeneracy in your fitting model.  Click "Apply" or "Apply and save" and rerun the Feff calculation.

Setting the fuzz to 0 will make the new pathfinder behave just like the pathfinder in Feff6.

Regards,
B

________________________________
From: ifeffit-bounces at millenia.cars.aps.anl.gov [ifeffit-bounces at millenia.cars.aps.anl.gov] on behalf of Simon Cassidy [simon.cassidy at stcatz.ox.ac.uk]
Sent: Wednesday, September 19, 2012 10:18 AM
To: ifeffit at millenia.cars.aps.anl.gov
Subject: [Ifeffit] path averaging in Artemis

Dear Bruce,

I've run into a problem with the new artemis' path finder. When running the attached feff.inp file in atoms, two separate paths are merged together.

The two paths it is merging are Se-Fe2 (N=4, reff=4.456) and Se-Fe3 (N=8, reff=4.465) which it gives as (N=12, reff=4.4621). The same problem occurs for these paths when starting with the Fe core.

The same feff.inp file generates the right path list when run in the old version of artemis. Please could you take a look at this?

Many thanks,

Simon


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://millenia.cars.aps.anl.gov/pipermail/ifeffit/attachments/20120919/ca97ef96/attachment.html>