Regarding standard deviation for ADP's

[04anuvra...@gmail.com]

Hello everyone, 

I am doing PDF analysis for my samples and need to report standard deviations (SD) obtained for the refined parameters. I have found out the standard deviation for all parameters except Uiso or Uequiv parameter. The SD is obtained for its components (u11, u22, u33,....) in the refinements itself but when exported to CIF format where we get Uiso or Uequiv we don't get any SD there. So, if anyone has any idea of getting the SD for Uiso or Uequiv please let me know it will be a big help for me. 

Thank you!

Regards,

Anuvrat Tripathi

[Simon Billinge]

Dear Anuvrat,

If you refined the variables that were associated with the ADPs then the reported SD will be in the results file.  In PDFgui ou can see this by selecting the fit and then the results tab.  The SD number will be associated with the VARIABLE number, but if, for example, you specified @5 for the ADP then look for the reifined result and the SD for variable 5.

On a side note, just because the program reports the SD, doesn't mean that it is right.  This depends on whether the uncertainties input into the program were right or not and on whether all the assumptions behind the mathematics of the propagation of statistical uncertaintiles are good assumptions.  This is discussed in more detail in the upcoming "PDF to the people" book.  The SD's reported by the program often seem to be quite helpful, even when those assumptions I mention above are not well followed, but use common sense rather than just reporting on numbers reported to you by a machine.....

S

--
You received this message because you are subscribed to the Google Groups "diffpy-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diffpy-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/diffpy-users/8941df98-21ac-42ab-8515-105207c2a159n%40googlegroups.com.

--

Simon Billinge

Professor, Columbia University