In addition to the problems mentioned by Bruce and Carlo, I see that number of independent data points is 13.8 and numbers of variable is 14. So you are really using the  "+1" factor in your favor here. For the fits to make physical sense, try to constrain your model such that the number of variables are about 2/3rd the number of independent data points (as mentioned by Scott on this mailing list earlier).

As far as the negative ss value is concerned, you can force it to be positive by setting the path parameter to return positive values only. Please look into Feffit documentation for the same.

Hope That Helps
Bhoopesh

On 2/23/07, hw26@njit.edu <hw26@njit.edu> wrote:
Dear all,

I am doing model fitting using Artemis. My sample is uranium solution with an
ionic liquid,  1-methoxyethyl-3-methyl imidazolium tetrafluoroborate
(MOEMIMBF4). The Fitting seems ok except that the ss_2 is negative. I don't
know how can I fix this problem. Can you get a chance to take a look? My
fitting is simple, only three single scattering pathes which were produced by
Six Pack. The fitting file is attached. I really appreciate it if you could
take a look and give some advice!

Best regards!
Hao


_______________________________________________
Ifeffit mailing list
Ifeffit@millenia.cars.aps.anl.gov
http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit