[Ifeffit] Artemis autoparams bug

[Bruce Ravel]

Scott,

There is certainly a problem in some situations that behaves as you
describe, but the situation is a bit different than you think.  A
while back, I made the mistake you point out in of Artemis'
configuration files.  This was fixed in a subsequent update.  

I just looked at the most recent update and its artemis.config file
is correct.  However, the way the configuration system works, anyone
who used artemis during its broken period has the broken autoparams
values in his personal configuration file.

In short, it's not really broken, except sometimes it is ;-)

One quick solution is to click the "Set all parameters to Artemis's
default" button on the preferences page.  This fixes the autoparams
roblem at the expense of undooing all you other customizations.

>From my (admittedly extremely biased) persepctive, this is one of
those situations where the project suffers from never having had any
volunteers to help with Windows packaging or with formal testing.  I
fixed the problem, but I am not sure how to propagate the fix in a way
that doesn't alter the behavior of the configuration system.

B

P.S.  I guess I should explain that the system-wide initialization
file is generated from athena.config at the time that the windows
executables are built.  The personal initilization file is regenerated
everytime you run artmis and contains the junction of the system and
personal init files.  The system file is read first and the personal
is read second, with values from the personal over-riding values from
the system.  That explains how a configuration bug from one update can
continue propagating through later updates.



On Tuesday 20 March 2007 20:41, Scott Calvin wrote:
> Hi Bruce,
>
> In response to some users, I was trying to update the ZnO tutorial so
> that it works with current versions of Artemis. Unfortunately, there
> seems to be a bug in the current version of Artemis. I'm posting it
> here on the mailing list because although there are very easy
> workarounds it's likely to give some newcomers trouble.
>
> I have Artemis 0.8.007 on Windows XP.
>
> To see the bug, make sure do_autoparams is on. Also make sure you're
> working with a new project.
>
> Now create and run a feff calculation in the usual way.
>
> When I do this, the guessed parameters created are:
>
> amp for S02
> delr for delE0
> dr for delr
> ss for sigma^2
>
> The second of those is quite confusing, and the third doesn't work.
>
> Of course, the workarounds are simple (e.g. don't use autoparams or
> change the names to something that works). But, since Artemis
> defaults to autoparams on, it seems as if this could be off-putting
> to newcomers.
>
> --Scott Calvin
> Sarah Lawrence College
>
> _______________________________________________
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

-- 
 Bruce Ravel  ---------------------------------------------- bravel at anl.gov

 Molecular Environmental Science Group, Building 203, Room E-165
 MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007

 Argonne National Laboratory         phone and voice mail: (1) 630 252 5033
 Argonne IL 60439, USA                                fax: (1) 630 252 9793

 My homepage:    http://cars9.uchicago.edu/~ravel 
 EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/