Create a Forum Topic by Logging In or Creating an Account

enquiry: zero standard errors for the XRR fit parameters

Wed, 01/20/2021 - 11:49 pm

I have done X-Ray Reflectivity Fitting using Motofit package in IgorPro 6 environment. I am getting standard errors of the fit parameters/coefficients to be zero. I have attached a screenshot, where you can see all the fitting parameters in the bottom window are having zero standard deviations/error bars. How to get finite standard errors? If anyone can help, please do let me know.

X-Ray Reflectivity fit using Motofit in Igor Pro 6

ilavsky

Did you try to contact Andy (Andrew Nelson, Ansto [last time I had contact with him]) author of the package? He wrote the package as well as Genetic optimization xop you are using. Based on help file included with Genetic optimization xop you should get error estimates. Not sure why you are not.

Here is good starting point : http://motofit.sourceforge.net/wiki/index.php/Main_Page

IMHO: That is way too many parameters for featureless curve. I do not believe these results can make any sense, it is possible those 0 are telling you that it fit should not be trusted.

January 22, 2021 at 06:23 pm - Permalink